ee e Well control: principles and practices e e Tew Taining ag v4.00Topic 1.1 Introduction During the majority of operations associated with drilling, completing, working over and eventually abandoning a well it is necessary to maintain control over the fluids that occur in the pore-spaces of formations being pene- trated by the well. These fluids can be subject to extreme pressures and temperatures in-situ although these are not pre-requisites for the fluids to cause well control problems! Failure to maintain control over these fluids can result in a spontaneous and sometimes rapid flow of the fluid into the well bore. The rate of flow is deter- mined by the degree of imbalanee between the well bore and reservoir pres- sures combined with the permeability of the reservoir. In its initial stages, such a flow is called a kiek. When such a flow is not controlled and deterio- rates in an uncontrolled manner it is described as a blow-out. Blow-outs can have a very visible environmental impact and, for that reason alone are very damaging for the Operator. The initial stages of a blow-out can also be very hazardous to personnel and cause major damage to equipment in the vicinity of the well. Control and recovery costs can be in the order of $10- $100 million. However, the blow-out can also cause significant damage to the producing reservoir through depletion and creation of preferential gas and water flow paths. It can also have a secondary impact on overlying forma- tions which may become polluted or abnormally pressurised. These factors impact on operations long after the surface environmental impact has been. resolved. It is, therefore. critical that Well Engineering staff know how to manage this hazard through * prevention - using primary control techniques and * control and recovery - if an under-balanced situation does occur; how to control it and regain primary control. The procedures associated with regaining primary control are called secondary control measures. These aim to regain control with minimum impact to the immediate and long term integrity and productivity of the well. Should these measures fail then more drastic tertiary well control measures may be applied. IWCF Training a Page 7 v1.00Topic 1.2 Formation pressures 1.2.1 ORIGINS ‘Two values are significant during the planning and drilling of a well; the pore pressure gradient and the overburden pressure gradient. To understand these gradients and the pressures that they can generate and therefore be encountered whilst drilling it is necessary to appreciate the origins of the pressures. The majority of the formations drilled through in hydrocarbon exploration and production operations is sedimentary. The material has settled out either in air or water and is subsequently covered by more material and buried. The majority of this action ocours in water. Depending on the nature of the material it behaves differently when compressed. Fine clay particles tend to have little inherent strength and so deform and compress tightly together, squeezing out the water around it to form claystones and shales with little or no porosity or permeability. Coarser and stronger, sand parti- cles tend to retain some strength and support each other leaving gaps between, called pores, which will generally be initially filled with water. Over time, minerals in the water that they are deposited in can precipitate and cause the grains to become cemented together. Limestone, originating from calcareous material in the water, tends to lose porosity under pressure and temperature. However, its friable nature often causes it to crack or fracture over time and these then can become filled with water or other fluids. Consequently, under ideal circumstances, a stratigraphic column of rock formation consists of two phases, a solid and a fluid phase. The solid phase is formed by the rock material, and the fluid phase by water, oil, or gas which fills any pore or fracture space between the grains. The degree to which these pores or fractures are linked such that fluid can flow through the rock is the permeability. The total weight of any volume of porous rock (W1) equals the sum of the weight of the grains or matrix (Wm) and the weight of the pore contents (Wo). W= Wn + Wo Considering a column of unit cross-sectional area from surface to a given depth gives the vertical load per unit area at that depth, ie. the vertical stress or pressure in the formation. W, yields what is known as the over- burden stress denoted by S; W,, yields the vertical matrix stress oy and W, Page 6 : IWCF Training v1.00yields the fluid pressure or Py S =o, +P, ‘The rates of change of these stresses with depth are called gradients. in units of psi/ft or kPa/m, which are just a different way of expressing the density of the material. Taking the depth as Z, which is conventional practice: the bulk density P= x = s the matrix density pee a doy 3% 70 105 Pressure im MPa i 5 i 19.4 kPaym (0.86 pstity = Depth x 1000 m © & Depth x 1000 1 20.6 kPaim (0.91 psutty 21.4 kPalm (0.95 pslit) - “ggostatie® or *overburcent L as 22.3 kPalm (0.99 paut) “under-pressures* sover-pressures* hydrostatic 1900 psi Figure 1.1 : Pressure gradient plots A convenient concept is the equivalent gradient which is the total stress (or pressure) divided by the total depth. The convenience comes because the only way in which pressure can be applied in a bore-hole is to fill it with a liquid of certain density and, if necessary, apply additional pressure at the top of the column. The equivalent gradient at any depth is the gradient of the liquid column which would produce the same stress/pressure as exists in the forma- tion without the application of additional pressure. IWGF Training 7 Page 9 v1.00‘The relationship between pressure and depth can be illustrated using a graph. Custom is such that these graphs are plotted with pressure (P) on the horizontal axis and depth (Z) on the vertical axis, with zero (= datum level) at the top corresponding to the physical system. A constant gradient is plotted as a straight line (see Figure 1.1). For pressure and gradient calculations during day-to-day operations the datum level is usually derrick floor elevation (dfe), since that is the top of the column of drilling fluid (in some areas the top of the kelly bushing is taken as datum but in practice the difference is negligible). Pressure plots are also frequently used to compare data from different wells. For this purpose a common datum plane has to be defined. The datum level is usually mean sea level for offshore wells; for land wells it can be any convenient level such as the average elevation of the area above mean sea level (rounded off to a convenient figure). In the latter case the datum level has usually been defined during the seismic campaign. ‘The overburden gradient or bulk density of formations penetrated can be measured by the bulk density logging tool. When these densities, measured in a well, are plotted, a step-like function of the bulk densities versus depth is obtained. Each step represents a change in lithology. Integrating the bulk density measurements to a given depth gives the overburden stress at that depth, and averaging these in an area will give a general trend like the curved line in Figure 1.2b. If no existing well data exists it is possible to make an estimate of the over- burden gradient. Assuming formation rock has a density of approximately 2650 kg/m? (26 kPa/m or 1.15 psi/ft): A clay formation near the sediment/water interface has a porosity of say 75%. Therefore the gradient of the bulk clay/water mixture is: 0-25 x 26 kPa/m + 0-75 x 9-81 kPa/m = 13-85 kPa/m or 0-25 x 1-15 psi/ft + 0-75 x 0-433 psi/ft = 0-612 psi/ft A sand formation near the sediment/water interface has a porosity of 35%. The combined bulk gradient of the mixture in this case is: 0-65 x 26 kPa/m + 0-35 x 9-81 kPa/m = 20-33 kPa/m or 0-65 x 1-15 psi/ft + 0-35 x 0-433 psi/ft = 0-899 psi/ft The overburden gradient line (Figure 2.1.2a) is, in actual fact, a curved line as compaction normally causes the density of rock to increase with depth. This behaviour depends on rock composition and porosity. As a rule of thumb the overburden gradient is usually taken to be 22-6 kPa/m (1-0 psi/ft although it can vary between 13-8 kPa/m (0-6 psi/ft) and 24-86 kPa/m (1'1 psi/ft). In areas where active sedimentation takes place like the Gulf Coast, offshore Page10 IWGF Training v1.00North West Borneo and Nigeria a value of 20-6 kPa/m (0-9 psi/ft) is more accurate due to the amount of water retained in the rock. Formation pore fluid density can be obtained by using logging tools which measure the pressure at a series of depths within the formation, Measured water gradients are usually between 9-81 kPa/m (0-433 psi/ft) for pure fresh water and 10-52 kPa/m (0-465 psi/ft) but may be as high as 11-31 kPa/m (0-5 psi/ft) for saturated salt water pore fluid. Sea water is assumed to have a gradient of 10-07 kPa/m (0-445 psi/ft). 1.8 1.9 2.0 24 2.2910 10 6200 «304050 GOP overburden depth in 1000 feet wetcsnn | ee 1" u 0.800 0.900 = 1.0000 2.000 4.000 6.000 8.000 10.000 psi density log ‘8. As measured by the bulk b. The overburden stress gradient doneity tool in gloe Figure 1.2 : The relationship between bulk density and overburden gradient 1.2.2 PORE PRESSURE PROFILES Formations are conventionally classified as normally pressured, under-pres- sured or over-pressured. Any pore pressure of the liquid or gas phase in the formation, as measured by logging tools, may be plotted on a graph similar to the one shown in Figure 2.1.1, and as indicated on that graph the position in which it is plotted will indicate into which pressure regime it falls. These pressure regimes do not have strictly quantitative definitions, being based on the practical matter of what drilling fluid gradient is required to balance the pore pressure. Because of this the boundaries between them are Jess well defined than shown in the graph and may vary according to the context and to local custom. IWCF Training Page 11 Vv 1.001.2.2.1 NORMAL PRESSURE REGIME Looking at Figure 1.1 it can be seen that two cases of hydrostatic gradient are shown; these correspond to water with zero salinity and water with a highish (but not extreme) salinity. In general if a plotted pressure falls between these two lines the formation is said to be normally pressured at that depth and a normal unweighted drilling fluid can be used while drilling through it. A normally pressured formation will usually have a hydraulic connection to the water table level although this does not have to be vertical, i.e. the connection could happen some distance away via connections with other formations, non-sealing faults ete. In this case the water table will generally be at approximately the same elevation as the drilling location. Ina region known to have unusually high formation water salinities, and thus densities, the formation pressure would still be called “normal” or “hydrostatic” even though the plotted pressure is in the overpressure region of the chart (as long as it is consistent with the known pore fluid densities). 1.2.2.2 SUB-NORMAL PRESSURE REGIME Formations with pore pressures which plot below the hydrostatic region shown in Figure 1.1 (or below the locally accepted normal pressure region) are said to have a Sub-Normal or Sub-Hydrostatic pressure regime. If a formation in the region mentioned above has a pore pressure equivalent to a gradient of 10.2 kPa/m (0.45 psi/ft) it would probably be referred to as under- pressured, depending on local custom, even though it would be a normal pressure elsewhere. Sub-normal pressures can be caused by the following situations: * Low water table or high drilling location elevation * Reservoir depletion * Tectonic extension after compression 1.2.2.2.1 Low water table or High Elevation In mountainous or arid areas it is possible for the water table to be deep below the surface at the drilling location. The top hole is then drilled through dry rock. Here the pore pressure is zero and the matrix stress alone defines the overburden down to the water table. Below the water table p, then plays a part. This situation is illustrated in Figure 1.3. For example if the water table is at 1000 m (3,281 ft) bdf and the formation fluid gradient is 10 kPa/m (0-442 psi/ft) then the pressure P at a depth D will be given by: P= 10 kPa/m x (D - 1,000) m P= 0-442 psi/ft x (D - 3,281) ft From surface it is said that the formation pressure is underpressured by 1,000 m x 10 kPa/m = 10,000 kPa or 3,281 ft x 0-442 psi/ft = 1,450 psi. Page 12 IWCF Training V 1.00well elevation Figure 1.3 : The effect of a low water table or high location elevation 1.2.2.2.2 Depleted Reservoirs It has been stated that a normally pressured reservoir usually has a hydraulic connection to the local water table level. This connection can be tenuous and of low permeability, and can sometimes even have been lost as a result of sealing faults, salt intrusions ete. This situation is often described by saying that a reservoir has little or no natural water drive. Consequently, if such a reservoir is produced the pressure in the reservoir will drop by the process of expansion to less than that calculated from the normal gradient (unless it is artificially maintained by gas or water injection). 1.2.2.2.3 Tectonic extension after compression Ifa reservoir rock with permeability and porosity becomes sealed after depo- sition, and is then extended or expanded due to tectonic stresses acting on it, the effect can be to expand the fluid in the pore spaces and so reduce the pressure. The effect will be to reduce the pore pressure to less than that, generated by the local hydrostatic pressure gradient. 1.2.2.3 OVER-PRESSURED PRESSURE REGIME Formations with pore pressures which plot above the hydrostatic region shown in Figure 1.1 (or above the locally accepted normal pressure region) are said to have an Abnormal or Over-Pressured pressure regime. Over-pressures may be the result of: «High water table or low elevation + Hydrocarbon bearing formations * Abnormal sedimentary burial circumstances * Osmotic action Whilst sub-normal or sub-hydrostatic formations can cause drilling problems such as losses or stuck pipe due to the pressure difference between the well bore and the formation pores (assuming that the well is kept full with water as a minimum!), over-pressures can cause more significant problems such as borehole instability, kicks and blow-outs. Consequently it is important to understand how over-pressures can occur and how they can be predicted and identified at an early stage. IWCF Training ; i EEE Pegs £2. v1.001.2.2.3.1 High Water Table / Low Elevation This is the opposite situation to a low water table/high elevation described above. When the well is drilled from a location below the water table or into a permeable formation linked to an aquifer at a higher level, the formation pressure will be greater than that generated by a hydrostatic gradient from the rig floor. This situation can lead to artesian flow from shallow aquifers (as used in London and Paris for drinking water production). Typically, problems can occur when entering the first porous, permeable formation after drilling through surface clays, claystones or shales. 1.2.2.3.2. Hydrocarbon Bearing Formations The permeability of shale to hydrocarbons is extremely low and can be consid- ered to be zero. Once hydrocarbons are trapped below a sealing shale the system is closed at the top. Figure 1.4 shows the basic principle. As the aquifer is open to the base, the water bearing part of the reservoir will be hydrostatically pressurised. Although a normal hydrostatic pressure may exist at the hydrocarbon-water contact, the column of low density oil, condensate or gas on top will result in pore pressures above hydrostatic. The difference between the pore pressure and the hydrostatic pressure increases with distance from the Oil-Water Contact (OWC) ot Gas-Water Contact (GWC) with the highest over- pressure at the top of the reservoir (see Figure 1.4). This abnormal pressure resulting from displacement of water by less dense fluids or gas is called genetic pressure. Often there is a small pressure difference at the OWC as oil is pushed down through the pore throats replacing the formation water. This pressure difference can be substantial if pore throats are very small (eg. in chalk). The pressure is a result of the gradient required to produce flow and of working against the capillary effect. Figure 1.4 : ‘Gas effect” over-pressures 1.2.2.3.3 Abnormal Burial Conditions. The conditions under which the sedimentary material is deposited and subse- quently buried can give rise to abnormally pressurised formation. A number of these mechanisms are reviewed below. Page 14 7 WCF Training v1.00Undercompaction Abnormal pressures arising from rapid burial of water saturated sediments are referred to as geopressures (sometimes referred to as depopressures or undercompaction pressures). This phenomenon is the mechanism that has caused the majority of overpressured formations in existence. Depo-, geo- or undercompaction pressures are created by the resistance of low permeability rock to the escape of trapped fluid. The permeability of clean sandstones, limestones and dolomites is generally too high to develop overpressures from undercompaction. However, the permeability of clay or shale decreases with compaction to extremely low values. As the process of burial proceeds a specific increment of overburden load normally causes a rock matrix to compact by a specific corresponding amount, provided that there is enough time available for the pore fluid to escape from the reducing pore space. Under such conditions the pore prossure in the rock remains in hydrostatic equilibrium with that in the overlying formations and remains normal. If however, the burial process proceeds too quickly the pore fluid will not be able to flow out of the formation at a rate which allows the same matrix compression rate to be achieved. The fluid will then have to support part of the overburden load and will therefore become overpressured. The formation is said to be undereompacted. An alternative way of looking at the situation is that as the rock is put under compressive stress by the increasing overburden the pore fluid always supports part of this increasing load. The increasing pore pressure thus creates a differential which causes the pore fluid to flow out of the formation. If the ratio of permeability to rate of increase of load is high enough the system will stabilise with a low elevation of the pore pressure. If the converse is the case equilibrium will be reached with a high pore pressure. Significant factors in creating an overpressure under these conditions are: * the permeability of the clay or shale which decreases with compaction. * the thickness of the shale through which the water must escape. ‘+ the potential compaction that could occur in the shale at the applied load. + the rate of burial. As mentioned above continuous beds of sandstones, limestones and dolomites do not become undercompacted because their permeability is high. If, however, a lens of a high permeability material is eompletely enclosed in a compacting clay (later to become shale) the pore fluid can no more escape from that than from the clay itself. The pore fluid will thus remain in pressure equilibrium with the fluid in the clay immediately surrounding it and the lens will become undercompacted and overpressured Due to the minimal permeability of shales, drilling into an undercompacted and therefore overpressured shale will normally only lead to borehole IWGF Training 7 Page 15 v1.00stability problems. Drilling into an overpressured porous lens can lead to a kick. However, if the lens is only water filled even a very small expansion of the sealed system will result in a substantial drop in fluid pressure, given the low compressibility of the pore water (4-35 x 107 kPa! or 3 x 10° psi). If, for example, the pore space is allowed to expand by as little as 0-1 of a per cent, the pressure of the pore water will decrease by 0-001/4-35 x 10-7 kPa-l = 2,299 kPa (0-001/3 x 10° psi-! = 333 psi). If, on the other hand, the formation has become partially gas filled due to degradation of material in the surrounding shale, the potential exists for a high pressure, high volume gas kick. Diagenesis Phase changes in the matrix material of the rock can cause overpressures. Some types of rock under the influence of pressure and temperature alters their structure and composition, resulting in a reduction in pore volume and/or an increase in the amount of formation fluid. Such a change could cause a pore pressure increase in a sealed system. For example in sand/shale sequences montmorillonite or smectite clays change to illite at about 220°R. This alteration frees water with a volume in excess of the reduction in volume of montmorillonite or smectite altered. If expulsion is inhibited, these liquids will, with continued burial, absorb increasing load. The effective stress will increase more slowly and abnormal pore pressures will result, ‘Tectonic pressures Tectonic phenomena such as sliding, piercements (volcanic), shearing, diaperic movements (salt, clay), uplifts (mountain building), thrust, ete. can add energy to the pressure system in a geological area resulting in abnormal Pore pressures in formations from which fluids cannot escape. Osmosis ‘Two reservoirs separated by an impermeable membrane may develop a pressure differential across the membrane as a function of a difference in the salinity of the formation liquid. As clay may act as such a membrane and because salinity changes occur in the subsurface, osmosis could have some potential to create abnormal pressures. Tests have proved that a pressure differential of 21 to 28 kPa (or 3 to 4 psi) could be generated by osmosis across a clay membrane. 1.2.3 PREDICTION Due to the drilling problems that both abnormally and sub-normally pres- sured formations can cause it is important to be able to predict them. Depending on the nature of the drilling operation this may be during the planning stages of a well or whilst actually drilling the well. Page 16 IGF Training V 1.001.2.3.1. PREDICTION WHILST PLANNING 1.2.3.1.1 Offset Well Information When planning a development or appraisal well, information from offset wells can provide indicators of over-pressure. Key information is: * Pore Pressure measurements (e.g. RFT logs) * Drill-speed logs - drillability should reduce with depth in shales (see later section on “d” exponent) * Formation bulk density logs (compaction and thus bulk density should generally increase with depth in shales) Formation Fluid contact levels (OWC, GOC, GWO) - these tend to remain almost constant across a field (assuming the absence of sealing faults) and thus their depths, combined with formation depth estimates, can provide sufficient information to estimate formation pressure gradi- ents along the proposed well path. Given that depth prognoses tend to have margins of error associated with them, it is normal to make pore pressure prediction estimates based on the worst reasonable case scenario (but not on the worst theoretically possible case). This is generally a combination of a reasonable estimate of the maximum reservoir height combined with a reasonable estimate of the maximum hydrostatic pressure. ‘This data can be used to make a good estimate of the pore pressures that may be encountered in the new well and so the easing scheme and drilling fluid program can be designed accordingly. However, pressure data has to be checked carefully to ensure that field production operations have not caused changes in the reservoir characteristics since the source well was drilled. These issues will be discussed further later in the Part 1.2,3.1.2 Regional Geology Because the majority of causes of overpressures are related to the deposi- tional environment or history of the sedimentary rock, the prediction of over- pressures during well planning, in the absence of any offset well information (eg. when drilling a wildeat or first well in an area), relies on either looking for indications of sedimentary activity that has the potential to generate over- pressures or, alternatively, extrapolating symptoms of overpressure that are visible at or near the surface to underlying formations. In common with hydrocarbons, often the only way to determine if overpressures actually exist is to drill! Complex studies of stratigraphic and hydrodynamic features have been conducted in some areas of the world. These can be useful to identify shallow overpressures caused by artesian flow conditions ete. but are less likely to yield information about overpressures in hydrocarbon bearing formations. Studies into the depositional environment of an area may indicate the type of rapid burial that can lead to over-pressures. Mathematical modelling is being applied to the action of sedimentation and burial with the objective of IWCF Training 7 _____ Page 17 v1.00simulating the process of compaction and so identify areas where overpres- sures may exist. 1.2.3.1.3 Geophysics It is rare that exploration or appraisal drilling is initiated prior to the comple- tion of seismic profiling of the area. With modern processing techniques it is possible to identify characteristics in the survey that suggest the existence of over-pressured formations. Primarily they identify undercompacted material due to poor response or slow transit times. These techniques become more reliable when combined with log data from wells in the same province. Actual sonic transit times in these wells can be compared to seismic data to identify where undercompaction appears to have occurred. Seismic surveys can show the presence of geological features known to cause overpressures such as salt domes (causing formation distortions), growth faults (in areas of rapid burial) ete. In addition, the presence of gas can increasingly be suggested by seismic data; bright spots indicating high ampli- tude reflections from gas filled (and so relatively low bulk density) formations being identified on the seismic maps. 1.2.3.2 WARNING SIGNS WHILST DRILLING It is normally possible to identify that the well is entering an area of over- pressure before the latter manifests itself by a kick from a porous and perme- able formation. To exist, an overpressured formation must be sealed, at least from above, by an impermeable layer. Consequently a pressure gradient will exist across the impermeable layer and the effects of this gradient provide the warning signs. The nature of the impermeable layer is important; a very thin but extremely impermeable layer (such as salt and limestone) will not provide as many clues as one where the pressure gradient permeates the formation (eg. shale). The formation where pressure moves from normally pressured is called the transition zone. 1.2.3.2.1. Drilling Parameters In general, the onset of abnormally pressured formations is associated with an increase in the drillability of the formation. This is a deviation from a general trend of reduced drillability with depth due to compaction, inereased stresses etc. Basically this increase can manifest itself in two ways. In shale, where this shale is a pressure transition zone, the increase will be gradual, while, when entering the overpressured reservoir itself the increase will be more sudden. The latter is normally called a drilling break. It should be noted that not all increases in the drilling rate are a result of an increase in formation pressure. It could simply have been the result of a ¢ in lithology. However, any unexplained increase in drilling rate should always be investigated. A number of attempts have been made to link the drilling parameters to generate a measure of drillability . The best known was by Jorden and Page 18 IWGF Training v1.00Shirley * who demonstrated empirically that in homogeneous shale forma- tions, and with all other factors being equal, penetration rate is proportional to the rotation rate of the bit. R=k,N where: R is the rate of penetration Nis the rotation rate ky is a constant of proportionality (that is only valid for one set of circumstances) Jorden and Shirley also showed empirically that the penetration rate depends on the weight on bit and hole size according to an exponential rela- tionship : wy! Rota x(S) where: W is the weight on bit D is the hole diameter kp is a constant of proportionality (that is, like k, , only valid for one set of circumstances) dis an exponent which, relating penetration rate to weight on bit per unit hole diameter, gives a measure of the "drillability” of the forma- tion It was shown, still empirically, that combining the two above equations into a one: = R=KxNx ¥) gives a good approximation to the behaviour of a real system in practice. Given that the utility of such a relationship is to compare the drillability of similar formations in similar circumstances, rather than to establish the absolute value of a well defined parameter, the equation is simplified by dropping the constant. The relative drillabilities of similar formations in which only the penetration rate, weight on bit, rotation rate and hole diameter vary is then given by the value of “d” in the equation wy R=N x( a in which case “d” is known as “the d exponent” and wee log We * Jorden, J.R. and Shirley, O.J., “Application of drilling performance data to overpressure detection,” Journal of Petroleum Technology, November 1966 IWGF Training Page 19 v1.00Given again that the objective is to make comparisons rather than establish absolute values, the units in which penetration rate, weight on bit, rotation rate and hole diameter are expressed do not affect the final result. When working in oilfield units, in which penetration rate is measured in f/hour, weight on bit in lbs, rotation rate in rpm and hole diameter in inches, the convention has grown to insert factors to convert diameter to feet, to make the time units consistent and to convert the weight on bit units into Ibs x108. Conventionally thus toe( a 60N log| 12W 10°D It has been found that the “d exponent” can additionally be adjusted for drilling fluid density variations. This gives the compensated d exponent, “d,” . normal pressure gradient density of drilling fluid in use Although the connection cannot be justified mathematically, the use of this compensated d (“d,”) exponent has proved advantageous. 4, =dx ‘These exponents were developed for application in homogenous shale inter- vals. They are less reliable in variable lithology, sandstone and limestone. dexponent—» However, it is very likely that a sealing cap rock is a shale so it does have relevance. Additionally, the equation was specifically developed for Gulf Coast condi- tions where drilling shale is the norm; it should therefore be used with caution in other areas. Depth Figure 15 illustrates the indication of change given by the “d exponent” Other algorithms have been developed that require significant, processing of raw data such as Geoservice’s Sigmalog, Baroid’s Log Normalised Drilling Rate (LNDR) and Anadrill’s Instantaneous Drilling Evaluation Log (IDEL) or A-exponent. These and 7 others may be provided as a standard or additional Mud- Figure 1.5 Logging service. ee 1.2.3.2.2 Drilling fluid parameters: A number of indications can be observed in the drilling fluid returns. These, again, have to be assessed in conjunction with the drilling parameters and the drilling environment. Relevant parameters include: Drilling fluid gas levels ‘The gas content of the drilling fluid can be measured crudely using stand- Page 20 IWCF Training V 1.00@ alone methane (CH,) meters or more sophisticated gas chromatographs (usually in Mud Logging units). ‘The following parameters may be useful: Background gas levels (very small quantities of gas contained in clay- stone/shale ) generally increase from ambient values when drilling into undercompacted shales. Gas shows (from the pores of gas-containing reservoir rock being drilled) will be apparent when a hydrocarbon reservoir is entered. The degree of the shows is dependant on the pore pressure of the rock. Excessive gas shows suggests a minimal overbalance on bottom. Trip | Conneetion gas may be observed due to slight reductions in bottom hole pressure due to a loss of the back pressure due to fluid friction in the annulus when the pumps are stopped plus a swabbing effect whilst picking up to make a connection or when tripping. Again, high readings tend to suggest that there is minimal overbalance on bottom. Gas composition from a gas chromatograph - a service offered by Mud Logging Units - can also provide indications of overpressures. The ratio of methane to ethane generally reduces as levels of ethane increase in transition zones or overpressured formations. QS levels - The presence of increasing levels of HgS in the drilling fluid whilst drilling evaporites can also be an indication of the onset of overpressures Drilling fluid density Reductions in drilling fluid density may be caused by increased gas content (see earlier) or increasing water content. In low permeability formations this may be observed in place of a higher volume kick. Drilling fluid temperature ‘Temperature gradients (41/4Z) in undercompacted formations tend to be greater than in normally compacted formations. Normally, the flow line temperature of the drilling fluid, under conditions of uniform cooling ete. should increase regularly with depth. If the increase deviates from the established pattern it can indicate the onset of over-pressures. This effect is, however, generally masked by changes in circulation rate, drilling fluid characteristics, tripping as well as riser cooling effects, particularly in deep water operations. Drilling fluid resistivity Associated with increased formation water content in the drilling fluid, this parameter can back up others such as density provided there is suffi- cient contrast in the salinity of the drilling fluid and formation water. Increased lovels of sulphur salts in the drilling fluid whilst drilling evapor- ites ean indicate over-pressures (see HS). IWGF Training H Page 21 v1.001.2.3.2.3 Cuttings analysis Careful analysis of cuttings samples can provide warning signs of impending overpressures: Shale density Shale density can be measured relatively easily at the well site by Mud- Loggers. In areas of undercompaction the density will not increase with depth as fast as in normally pressured formations and may even decrease. Plotted against depth this is a useful tool in areas of long shale sections. Cation Exchange Capacity (CEC) Shale is composed of four different clay minerals, kaolinite, illite, montmo- rillonite and smectite. The ratios of these minerals change with depth due to the transformation of smectite into illite with increased pressure and temperature, A simple test can be performed to estimate the amount of smectite in a sample of cuttings and this value, when plotted against depth should show a negative gradient. Deviation from this trend in a shale formation is indicative of entering a transition or overpressured zone. Cuttings shape, size and abundance In general, as a transition or over-pressured zone is penetrated, increased borehole instability caused by a reduced drilling fluid pressure overbalance can lead to increased volumes of larger, more angular cuttings. This may be associated with other indications of borehole instability such as over- pulls ete. It is important to realise, however, that such features are not indicative in isolation as they can be caused by other factors such as tectonic stress etc. 1.2.3.2.4 Wireline logs & formation pressure evaluation Many logging tools used for petrophysical analysis can give indications of overpressured formations or transition zones, predominantly through identi- fying shale undercompaction. Density Under normal compaction conditions, shale density increases with depth. Consequently, a change in the density gradient can suggest a transition zone or overpressures. Resistivity Shale density and therefore resistivity tends to increase with compaction and depth. Reductions in resistivity can indicate overpressures. Sonic Sonic transit times reduce as density increases. Thus, an increase in the transit time can indicate overpressures. Page 22 IWGF Training v1.00Formation pressure Ultimately, a series of formation pressure measurements will provide quantitative evidence of the existence of over-pressures. However, these require porous, permeable formations and cannot give warning of the exis- tence of a transition zone. 1.2.3.2.5 Well seismic evaluation Seismic check shots or a Vertical Seismic Profile can be conducted in the well and the time/depth data used to correct the original seismic interpreta- tions which may have been based on assumed velocities in various rock formations. These corrections can assist in more accurately predicting the depth of deeper sites of undercompaction and thus possible overpressures. 1.2.4 SHALLOW GAS 1.2.4.1 INTRODUCTION Shallow gas deposits account for 41% of all recorded blowouts in the Gulf of ‘Mexico and the North Sea since 1980. Over 50% of these blowouts resulted in major damage to the drilling unit and many have resulted in serious injury and death to personnel. Characteristics of shallow gas blowouts are very little warning, very high flow-rates, large volumes of abrasive sand, rapid equipment failure, gas leaks, ignition and fires, all in conjunction with exces- sive noise. If one commences it is virtually impossible to recover control until the gas flow stops naturally, either due to exhaustion of the reservoir or collapse of the bore hole. 1.2.4.2 DEFINITION Various definitions for shallow gas deposits exis EP 88-1000 SIPM Shallow Gas Procedure Guidance Manual: “gas that is encountered in the well which cannot be closed in, because well shut-in pressures combined with the hydrostatic head of the fluid in the well bore will probably result in formation breakdown and subsequently cratering of the well. If BOP protection and normal well killing proce- dures can be applied the gas is not considered to be shallow gas”. Shell Expro: Well Engineering Information System (1995) : Shallow gas shall be taken to mean: i. Hydrocarbon Gas accumulations encountered before setting surface casing (first casing on which a BOP is installed) ii, Hydrocarbon gas accumulations encountered after setting the surface casing but where the well cannot be closed in. IWC Training Page 23 V 1.001.2.4.3 ORIGINS OF SHALLOW GAS Shallow gas can derive from two sourees: Biogenic Gas Biogenic gas is formed at shallow depth from the decomposition of organic material (i.e. “swamp gas”). It is primarily methane and tends to be found in-situ in the decomposing material or in a very close reservoir. This type of shallow gas is particularly prevalent in swampy areas as well as in areas of rapid sedimentary deposition where there is a significant amount of organic material being deposited. Petrogenic Gas Petrogenic gas is generated at depth from the degradation of hydrocarbon source material at elevated temperature and pressure and has subse- quently migrated into a shallow reservoir rock overlain by a seal. ‘Migration can have occurred naturally over geological time or may have occurred recently due to poorly cemented casing strings, internal blow-outs ete. Alternatively, in rare circumstances, a deeply buried gas reservoir can become uplifted or exposed near surface by erosion of the original over- burden. Petrogenic gas is predominantly composed of Methane, Ethane and Propane. Gas of either origin will only form a reservoir if it exists or subsequently migrates into a permeable and porous formation overlain by a sealing forma- tion. The gas is commonly found in highly porous and permeable unconsoli- dated sands overlain by clay or claystone. Due to the depth of burial it is unusual for the reservoir to be overpressured however, the “gas effect” can cause the top of the reservoir to be abnormally pressured with respect to the local hydrostatic gradient. Shallow gas pockets can also form in cold environments without the presence of an impermeable formation. Migrating gas cools to form hydrates (solid mixtures of water and gas molecules) which, being impermeable, trap addi- tional gas. This environment is becoming increasingly significant in deep water drilling operations where sea bed temperatures can approach 0° C. 1.2.4.4 OPERATIONAL CONSTRAINTS Given the shallow depth of burial and the very low formation strength of overlying formations it is not normally possible to drill with sufficient drilling fluid gradient to control shallow gas formations that are overpressured due to the “gas effect”. In addition it is generally not possible to close in a well in the early stages of a shallow gas blowout without the risk of breaking down the formation at the previous casing shoe (if any) and cratering the formation around the well. Consequently the only well control method available is to divert the flow away from the rig whilst attempting to regain control of the well or evacuating the location. ‘The entrained sand and other material Page 24 IWCF Training V 1.00causes significant damage to the diverting equipment and its subsequent failure has been the primary cause of major shallow gas incidents, many trag- ically with fatal consequences. ‘The plume of a flow of shallow gas has been postulated to cause stability problems for floating drilling vessels, particularly drill-ships. The flow of gas. can also ignite at surface forming a hazard to the rig itself as well as to evac- uating personnel. Other characteristics of a shallow gas kick/blowout are: * difficulty in identifying initial well flow - particularly when drilling from floating units and there is no return of drilling fluid to surface. * a rapid increase in flow - as the well starts to flow additional gas rapidly reduces the hydrostatic head in the well and so increases the flow rate. Total unloading of the well to gas can happen in less than a minute leaving very little time for analysis and reaction and making controlled reaction to a shallow gas kick extremely difficult. If returns are to surface the noise levels are exceedingly high, adding to the diffi- culty of carrying out emergency procedures. + whilst control may be held whilst drilling, rapid flow can start when circulation ceases or when the hole is circulated clean of drilled cuttings. Swabbing whilst tripping can initiate a shallow gas kick and is difficult to identify when there are no returns to surface. For this reason the standard procedure is to pump out of the hole if there is any possibility of shallow gas. 1.2.4.5 PREDICTION The primary defence against shallow gas is its identification and avoidance at the well planning stage. As mentioned earlier, certain depositional environ- ments are more prone to shallow gas, particularly of the Diagenic origin. However, Petrogenic shallow gas has the potential to exist wherever a shallow sealing formation exists. In addition, deep water hydrate formations can exist even in the absence of a seal. The principle method of identification is through the use of Seismic Surveys. These can be either analogue or digital Shallow seismic surveys. Analogue surveys, combined with bathymetric measurements and a sea bed sonar survey, provide a very limited penetration into the sea bed, being primarily of use for selecting a suitable drilling location. Digital surveys can be processed and analysed to identify bright-spots in formations down to around 3,000 ft (1,000 m) and with thicknesses in excess of 15 ft (5 m)which are indicative of the presence of shallow deposits of gas. (refer to EP 88-1000 for additional data on shallow seismic surveys). Shallow seismic can be augmented by small diameter soil boring surveys (usually limited to 50 m penetration) and pilot hole drilling from a conven- tional drilling rig. Pilot holes can be logged using conventional petrophysical IWGF Training zs Page 25 v1.00 @tools (Induction / SP / Sonic / Gamma Ray / Caliper) to identify the presence of shallow gas deposits and evaluate its characteristics. Additional information on the operational precautions to be taken if shallow gas is present can be found in Topic 1.4.3.3. Page 26 IWGF Training V 1.00Topic 1.3 Formation strength 1.3.1. INTRODUCTION In Topic 1.2 the origins and potential for abnormally-pressured formations was explained. The basic principle of controlling these pressures , as discussed in Topic 1.4, is to balance or exceed these pore pressures with the hydrostatic head of the drilling fluid. This principle defines the minimum drilling fluid gradient that can be permitted (i.e. so that the formation pore pressure is just equalled or exceeded. However, another limitation exists on selection of the drilling fluid gradient; the maximum drilling fluid gradient that the formation can with- stand before fractures occur and develop in the rock surrounding the well bore, the formation breakdown gradient. These fractures have a tendency to propagate at a slightly lower pressure, the fracture propagation pressure, than that required to initiate them. The propagation opens up an increasingly large volume of space causing losses of the drilling fluid into the formation. It is also assumed that once a fracture has occurred in a formation then the formation strength is reduced to a value approaching the fracture propagation pressure. The relationship between these pressures is discussed in this Topic. ‘The formation being drilled may dictate an even lower maximum drilling fluid gradient. * In some cases underpressured formations can be encountered such that a significant differential pressure exists between the drilling fluid column and pore pressure. * In some cases fractures and cavities are encountered which cannot be sealed off by a filter cake built up by the drilling fluid, even with the addition of very coarse lost cireulation material (LCM), and the drilling fluid itself is then lost to the formation unless its gradient can be reduced to the formation gradient. Even if this can be done so that no losses occur in the static situation there will still be losses while drilling and/or circu- lating because of the increased fluid pressure on bottom due to friction losses in the annulus and the cuttings load in the annulus. © In some areas both of these situations occur simultaneously. In that case even a column of water cannot be supported and drilling blind (without fluid returns) until an impermeable zone has been reached may be neces- sary. Once the loss zone has been passed and an impermeable zone has been reached a casing string has to be set to improve the bore hole strength before it is safe to drill deeper. The procedure of drilling blind requires a good knowledge of the forma- tions to be penetrated, and is not advisable in exploration wells. IWGF Training Page 27 v1.00Information about the anticipated formation strengths are required for the preliminary casing design during the planning process. Where this data is based on empirical estimates or extrapolation of data from offset wells it is sometimes necessary to conduct a Formation Strength Test (see later) to check the validity of the information and so the integrity of the casing design. More information on the subject of formation strength and its measurement is to be found in the following EP Reports: * EP93.1993 The KSEPL Rock Mechanics Manual, Part 1 - Introduction and Basics by CAM. Veeken (Report RKMR. 93.052, ). * EP922000 Casing Design Guide 1.3.2 FORMATION STRESSES One of the principle influences on the strength of a particular formation is the stress field existing in the rock prior to drilling through it. It is important to understand this phenomena and the effects of drilling operations on it. Any element within a confined solid body can be assumed to have forces acting on it which can be resolved into three components, at right angles to each other. In any particular case oy Se the values of the components will depend on a the direction in which it is chosen to resolve the 6. forces. If the directions in which the forces are resolved are such that one component reaches a maximum and one reaches a minimum the stresses within the body are known as the prin- cipal stresses, normally known as 0}, 62 and Gg. : 0; is the maximum principal stress and o3 Figure 1.6 : Principal stresses is the minimum principal stress. oO G; 1.3.2.1 STRESS DISTRIBUTION IN THE FORMATION In the case of a geological formation one of the forces acting on any element within the rock is the weight of the rock (plus pore contents) above it - known as the overburden. This gives the vertical stress o, . There are also hori- zontal confining forces Gy, and O49. In the majority of sedimentary formations the greatest stress is that caused by the overburden, and ay is 0}. It follows that oy; is Gy and Op is 03. The two horizontal components, oj; and ojo, combine to form a hoop shaped stress distribution in the horizontal plane. The shape of this field is depen- dant on the ratio between the maximum and minimum horizontal confining stresses. In ideal circumstances these stresses will be equal (ie. a uniform horizontal stress) and the stress field will be circular. In rocks which have been subjected to folding and/or faulting and/or other tectonic stresses (which applies to the majority of hydrocarbon reservoirs, Page 28 IWGF Training v1.00otherwise they would not have formed traps) the horizontal stresses may not be equal, and in extreme cases one of the maximum horizontal stresses may be greater than the vertical stress. (In which case o,, would be 0},). However such cases are rare, and extreme cases are very rare, because rock formations are plastic in a geological time frame and after tectonic events any abnormally high horizontal stresses will relax and a close approximation to the usual stress situation will re-establish itself. 1.3.2.2 STRESS DISTORTION CAUSED BY THE BOREHOLE ‘The removal of formation to make a well bore distorts the pre-existing stress field in the rock. An approximate calculation of this distortion may be made by assuming that the rock is elastic, the bore hole smooth and cylindrical, and both the bore hole axis and the regional principal stress , are vertical. In general, none of these assumptions are precisely correct, but they will provide a close approximation to the actual stresses. The stresses will therefore all be viewed as the effective stresses carried by the rock in addition to a hydro- static fluid pressure “P,”, which exists, and is assumed equal, within the well bore as well as in the rock. A calculation of the stress distribution around such a vertical bore hole could be made from the solution in elastic theory for the stresses in an infinite plate containing a circular hole, with its axis perpendicular to the plate, which was first obtained by Kirsch, and is also given by Timoshenko, and by Miles and Topping. Such tangential stress is known as a “hoop stress field” (Figure 1.7). It can be seen (see Figure 1.9a) that in every case the stress concentrations are local and that the stresses rapidly approach the undisturbed regional stresses within a few hole diameters. For the case in which 6/03 is 1.0, the stress field has radial symmetry and the tangential stress at the walls of the hole is +2 03. ‘The vertical component of the stress is also distorted in the vicinity of the bore hole. The initial vertical stress is equal to the effective pressure of the overburden. However, the magnitude of this distortion is small in comparison Figure 1.7: The hoop stress around a vertical borehole(ay is vertical) IWCF Training “Page 29 v1.00with the concentrations of the horizontal stresses. In addition it rapidly disappears, with distance, away from the well bore. 1.3.23 THE EFFECT OF PRESSURE APPLIED TO THE BOREHOLE ‘The application within the bore hole of a non-pene- trating fluid pressure, in excess of the original fluid pressure, produces additional stresses. The tangential stresses due to an additional pressure AP in the well bore are shown in Figure 1.8. The stresses given are those caused by AP \ alone, and to obtain the T complete stress field it is ae t necessary to superimpose pO these stresses upon those caused by the pre-existing regional stresses. This is o = “AP illustrated in Figures 1.9b : ae and 1.9¢ in which a pressure Figure 1.8 : Stresses cause by AP within the well bore equal to 1.6 6 is applied to the well bore for the case in which op/og = 1.4. That pressure is just suffi- cient to reduce 0 to zero. In all cases when the ,/6 ratio is greater than 1, the vertical plane across which the stress first becomes zero as the well bore pressure is increased is the plane perpendicular to os, the least principal stress. 1.3.2.3.1_ Rupture Pressures (formation Breakdown Pressures) In sections of a well bore through consolidated rock with any significant length it is probable that quite a few rock layers have been intersected. As (a ©) ©) oe \ 4 { f3 { i f -1.885 262 3 at hee PoEEEE = ar £1609 HHHes 186, =1.605) on 40% ap +6: 4 4P (ot/65= 14) Figure 1.9: Superimposing fluid pressure stresses on the in-situ stresses round a well bore Page 30 IGF Training v1.00cementing between layers will be weak if non-existent. the tensile strength between the layers can be negligible. Each layer, however will have its own unique tensile strength depending upon its structure, compaction and cemen- tation. This tensile strength will exist both vertically and horizontally but may differ in value. Consequently, the tensile strength of the formation can vary considerably in an interbedded environment. The hydraulic pressure required to produce a vertical separation (horizontal fracture) in horizontal formations is that pressure required to equal or slightly exceed the sum of the vertical compressive stress and the weakest of any inter-bed tensile strength; to initiate a vertical fracture it is necessary to exceed the sum of the minimum horizontal stress and the least of the forma- tion's tensile strengths. As tectonic effects determine which of these stresses is the greatest they will also influence whether a fracture will occur horizon- tally, vertically or even inclined in the well. As the pressure is increased, the plane along which a fracture will commence will be that across which the compressive stress is first reduced below zero Ge. it becomes a tensile stress) to such an extent that it overcomes the tensile strength in that plane. In the case of a smooth cylindrical well bore in an isotropic formation, this will be a plane perpendicular to the least principal stress. In cases where the largest principal stress 6; is vertical, the down-the- hole pressure required to start a vertical fracture may vary from the least principal (= horizontal) stress to a value of twice that value, depending upon the op/og ratio. A feature of vertical and inclined fractures is that they have to overcome a significant tensile strength in the formation before they initiate. However, once a fracture has occurred, the tensile strength will have been reduced and subsequent loading of the well bore will open this pre-existing fracture at a lower pressure. For the same reason, the formation breakdown pressure of a formation in a naturally fractured environment will be lower than that in an unfractured environment. 1,3.2.3.2 Injection pressures Once a fracture has been started the fluid penetrates it and pressure is applied to the walls of the fracture. This reduces the stress concentration that previously existed in the vicinity of the well bore. The pressure AP required to hold the fracture open in case of a non-penetrating fluid is then equal to the component of the undistorted stress field normal to the plane of the fracture. A pressure only slightly greater than this will extend the fracture indefinitely provided it can be transmitted to the leading edge. This can be seen from an ideally elastic solid, as shown in Figure 1.10. The normal stresses across the plane of a fracture near its leading edge are shown for the case in which the applied pressure AP is slightly greater than the original undistorted stress field o3, The tensile stress, as a result of AP in the bore hole, near the edge of the WCF Training Page 31 v1.00fracture approaches an infinite magni- tude for a perfectly elastic material. For actual materials this stress will still be so large that a pressure AP, only slightly greater than 0g, will extend the fracture indefinitely. The minimum down-the-hole injection pressure, required to hold open and extend a fracture, is therefore slightly in excess of the original undistorted regional stress normal to the plane of the fracture. The actual injection pressure will in general be higher than this minimum because of friction losses along the fracture. Figure 1.10 : 1.3.2.3.3 Pressure behaviour during injection tests ‘Stresses in the vicinity of a crack in a stressed elastic material when the pressure acting on 7 the walls of the crack is slightly greater than tee ee the stress within the materal © non-penetrating fluids and for various values of the 6/03 ratio, shows that there are in general two possible types of down-the-hole pressure behaviour during fracturing. These are illustrated in Figure 1.11. During a fracture test the pressure AP is increased and in one case (case ‘a’) the breakdown pressure may be substantially higher than the injection pressure. This would probably correspond to a horizontal fracture from a relatively smooth well bore or to a vertical fracture under conditions in which 0, and 63 were nearly equal, again assuming that o; is vertical. In the second case (case ‘b’) there is no distinct pressure breakdown during the treat- ment, indicating that the pressure required to start the fracture is less than or equal to the injection pressure. This would correspond to a horizontal or vertical fracture starting from a pre-existing opening, or to a vertical fracture in a situation where o,/og was greater than 2.0. Fluid pressure diferance between Fluid proseure diference between the formation and the well bore the formation and the well bore ' ' i | Fld pressureinthe welboe | | uid pressure inthe wel bore g /, Fri pressure in the formation 8 | /, Fluid pressure in the formation 8 J g z 7 3 3 3 z ee = oon (a) ©) Figure 1.11 : Idealised diagrams of pressure behaviour when creating a fracture Page 32 IWGF Training Vv 1.001.3.3 PREDICTION Whilst it is difficult to estimate the effect of additional tectonic stresses it is possible to make a rough approximation of the field around a well bore in “ideal” circumstances. This approximation will provide an order of magnitude estimate. In-situ stress measurement (i.e. Mini-frac tests) are required to determine the local stress distribution and orientation with any accuracy. However, values from one well may be extrapolated with some reliability over substantial distances, particularly the horizontal stress values. The undisturbed effective vertical stress o, is equal to the total pressure of the overburden, S, less the original fluid pressure, P,, existing within the rocks prior to disturbances such as fluid withdrawals. In algebraic form: o, = S-PBy. Under conditions of incipient normal faulting, the least principal stress 0 will be horizontal and will have a value of approximately one-third the effective vertical stress, 6, or 6. Therefore o3= (S - P,)/3. Since the additional fluid pressure differential AP required to hold open and extend a fracture should be slightly larger than the least principal stress, then: AP = (S - P,)/3. However, the total injection pressure P is given by: P = AP + Po. Therefore: P= (S + 2P,V/3 Dividing by the depth, Z, then gives: P/Z = (S/Z + 2 PJ/Z)/3, which is the approximate expression for the minimum injection pressure required per unit of depth in an area of incipient normal faulting. ‘The value of S/Z. is approximately equal to 22-6 kPa/m (1.0 psi/ft) for normal sedimentary rocks in most areas. Under normal hydrostatic fluid pressure conditions, PZ is about 10-4 kPa/m (0-46 psi/ft). Substituting these values gives P/Z = 14-5 kPa/m (0-641 psi/ft), as the approximate minimum value that should be expected in areas with incipient normal faulting. In different areas the oy/o, ratio may have different values and so maximum drilling fluid density or minimum injection pressures could differ. In compressional tectonics o, may be three times the effective vertical stress Gy and 80 G2 (S - Ps) x 3. The additional fluid pressure AP required to hold open a fracture should equal the least principal stress, which now is the effective vertical stress: AP = oy. As P=AP+P,,P/Ze(o, + PoVZ=S/Z= 22-6 kPalm (1-0 psi/ft). which is just another way of saying that to hold a horizontal fracture open the pressure has to support the whole of the overburden. Formation breakdown gradients play an overriding role in maximum drilling fluid densities, in casing design, in drilling operations and in injection of liquids in the formation. IWF Training Page 33 v1.001.3.4 MEASUREMENT 1.3.4.1 PROCEDURE For the purposes of drilling operations it is sufficient to know the maximum pressure that may be applied to exposed formations without causing damage and thus risk the generation of fractures and associated losses etc. It is not normally relevant in which direction such a fracture would run and whether it is a vertical, horizontal or inclined fracture. The standard operational procedure to ascertain the maximum pressure it is a Formation Strength Test. Whilst this may take a number of forms the general procedure remains the same. The test is normally conducted after drilling out the shoe track of a recently cemented casing string. The cement filled pocket is drilled out along with a minimum of about 20 ft (6m) of new formation. The test is conducted using a low volume high pressure pump (i.e. the cementing pump) and calibrated pressure gauges over a variety of ranges. The drilling fluid system pumps and gauges are not sufficiently accurate enough to perform the operation. The proce- dure is: * Circulate and condition the drilling fluid to a consistent density * Pull back the bit into the casing shoe * Ensure the well is full. Close in the well using the BOP pipe rams around the drill pipe * Open the annuli between the current and previous casing strings and monitor for flow * Slowly pump down the drill string until surface pressure approaches ca. 100 psi (700 kPa). Carefully measure tank levels ete. Pump uniform increments of volume - 0-1 to 0-25 bbl (0-016 to 0-4 m3) then stop and wait 2 minutes for pressures to stabilise. For each inere- ment the following are noted: - cumulative volume pumped, - pressure immediately after pumping ceases (final pumping pressure) - static pressure after 2 minutes (final static pressure) * Plot the cumulative volume pumped against both the dynamic and static pressures on a graph. * Continue to pump incremental volumes until one of the following occurs: - apre-determined limit pressure has been reached - the static pressure line deviates from a straight line (ie. a linear relationship between pressure and volume pumped) * Bleed off pressure at surface via the annulus (to prevent bit plugging!) Page 34 IWCF Training v1.00and monitor the returns. Determine how much fluid has been lost to the formation. It should be noted that the objective of a Formation Strength Test is not to break down the formation and generate/propagate a fracture. The point at which the pressure/volume plot deviates from a linear relationship is called the leak off point. It should be taken as the last measured point on the straight line; no extrapolation should normally be performed that would yield an increased formation strength. The leak off point is sometimes also called the formation intake point. A formation strength test that is terminated when a leak off point is identi- fied is called a Leak Off Test. This type of test is commonly performed on Exploration and Appraisal wells below any casing shoe that will form part of a pressure vessel (i.e. have BOPs attached at surface). It may also be performed to gain information to aid in the design of future wells. A formation strength test that is terminated when a pre-determined pressure is reached is called a Limit Test. Usually performed on development wells, these also test the integrity of the cement bond around the casing shoe. They are sometimes omitted in the absence of mobile hydrocarbons or over-pres- sured formations in the interval to be drilled. On rare occasions the decision is taken to perform a Formation Breakdown, Test. The pressure is deliberately increased to the point where a fracture occurs, the Formation Breakdown Pressure (FBP). This is usually marked by a sudden drop in surface pressure. Continued pumping now Breakdown will be possible at a lower t pressure - the Fracture . ; Fracture propagation (FPP). If pumping is stopped the surface Gee cooure pressure will drop rapidly Xe and then stabilise. This final pressure equates to the Fracture Closure ‘Cumulative volume pumped —_| Shut-in period Pressure (FCP). The ; ; ; relationship of these Figure 1.12 : Typical formation breakdown test result pressures under ideal circumstances is illustrated in Figure 1.12. Propagation Pressure j 1.3.4.2 INTERPRETATION The majority of competent formations will display a linear relationship between pressure and volume (essentially reflecting an elastic deformation of the rock). Permeable formations may show some separation between the final pumping and final static pressures as slight fluid loss occurs to the formation. The fluid loss control characteristics normally take effect rapidly however, causing the two lines to be almost superimposed. IWGF Training "Page 35 v1.00Shallow, unconsolidated formations can, however, display plastic deformation almost from the start of the-test with continued fluid losses to the formation. This can make interpretation difficult (although the formation actually does have very little strength!) and an estimation can be made by identifying where the static pressure trend deviates from that of the final pumping pressure trend. The actual Formation Strength Gradient (FSG) is calculated as follows: Maximum pressure at shoe Land: = FSG Shoe depth(tvd-bdf) - Derrick floor elevation Maximum pressure at shoe - Head of seawater column Offshore: FSG= eee Shoe depth(tvd- bdf) - Sea bed depth(bdf) For well control calculation purposes the FSG is normally expressed as an Equivalent Maximum Mud Gradient (EMMG): Maximum pressure at shoe EMMG = ~2(8-a)xin P2 P, a original drilling fluid gradient gas cut drilling fluid gradient bottom hole pressure with p, fluid atmospheric pressure reduction in BHP In = natural logarithm Gas cut drilling fluid can also occur continuously while drilling as a result of gas which is present in the drilled cuttings. Increasing the drilling fluid density does not stop the gas cutting effect. To confirm that gas cutting is caused by the cuttings only, drilling should be stopped and the well should be circulated bottoms up. If gas cutting does not reduce or cease after bottoms up, a potential kick condition may exist and the drilling fluid density should be increased. Flow checks Unexplained changes in the readings on the Flow Show and PVT recorders should be investigated promptly. This usually takes the form of a flow check. Circulation/rotation is stopped and the bit pulled off bottom. The well is then checked for flow. If flow continues after the circulation has been stopped then Secondary Pressure Control measures are initiated. Sometimes the flow check may be replaced by closing in the well immediately, particu- larly where kick tolerance (the maximum size of influx that can be safely circulated out without breaking down the previous casing shoe) is low, or the Page 44 IWF Training V 1.00PVT or Flow-Show indications are dramatic. A drilling break is a distinct change in penctration rate which does not always appear as a sharp change, but may be a gradual increase over a fairly long interval. A decrease in penetration rate can also be seen as a drilling break, depending on the type of formation which is being drilled (e.g. drilling from salt into carbonate). When a drilling break is encountered, a flow check should be made to confirm that formation fluids are not entering the well bore. The duration of the flow check depends on several factors and should be defined locally. Following a negative flow check (no flow observed), normal drilling may be resumed. If the flow check is not conclusive, the hole should be circulated bottoms-up before deciding to drill ahead. Flow-line indications Akkick is generally recognised by an instantaneous increase in the rate of Grilling fluid returns and/or an associated gain in drilling fluid volume in the active tanks. However a flow line show, observed when gas cut or liquid cont- aminated drilling fluid is returning from the well, does not necessarily mean that the well is kicking. Flow line shows can also be: * drilled shows * swabbed shows * high pressure - low volume shows As the bit drills a porous, permeable formation, some of the fluids contained in the pores of the cuttings will enter the drilling fluid stream. This drilled show can be characterised by: + a constant elevated gas content in the drilling fluid, correlated with a change in the lithology and penetration rate. * no peak in gas eutting or contamination of the drilling fluid at bottoms up * no increase in the volume of the drilling fluid. However, if drilling rapidly through a very porous gas containing formation the gas will have expanded by the time that it approaches the surface, and a minor tank level rise may be seen. Gas eutting normally results in a negligible reduction in the hydrostatic bottom hole pressure. However, when the overbalance is small (usually in top-hole drilling operations), it is important to keep the reduction in density of the gas cut drilling fluid to a minimum in order to avoid the well becoming underbalanced. When a drilled show occurs, it is not necessary to increase the drilling fluid density. Degassing of the drilling fluid is, however, required to prevent its gas content increasing with each successive circulation. If unchecked, this could WEF Training : Pages v1.00result in a progressive reduction of the hydrostatic pressure at the bottom of the hole. Swabbed Shows result from the temporary reduction in hydrostatic head of the drilling fluid column, caused by pulling the drill string up while making connections, or during a short trip. Swabbed shows are characterised by: © gas cut or oil/vater contaminated drilling fluid at the flow line when bottoms up occurs * no immediate tank level rise (in the case of gas, a minor rise will occur later when the gas approaches the surface) A swabbed show provides a useful indicator that the drilling fluid overbalance is only just sufficient in static conditions, or that there may be a need to adjust the drilling fluid rheology. As mentioned previously High pressure - low volume shows occurs when drilling in highly pressured formations with low permeability, or after trips with those formations open. High pressure - low volume shows are charac- terised by: * as the bit penetrates the formation the reading of the drilling fluid gas content will increase steadily as the flow area increases, but once the formation has been completely traversed the gas readings will correlate inversely with the circulation rate, due to a constant inflow regardless of circulation rate. * gas cut or oil/Avater contaminated drilling fluid at the flow line when bottoms-up occurs - this is a special case of the above. * no immediate tank level rise (in the case of gas, a minor rise will occur later when the gas approaches the surface) Hole stability problems (e.g. sloughing, torque, etc.) may also give an associ- ated indication of this pressure imbalance. Records Accurate records of all operations must be maintained, particularly of any unusual occurrences, These should describe all flow line shows, including the duration of the occurrence, drilling fluid return gradients, chloride content, etc. 1.4.3.1.2 Lost Circulation ‘The ability to circulate drilling fluid around the well is key to keeping the well under control and being able to drill ahead. Sometimes a situation known as lost circulation occurs when some or all of the drilling fluid which is pumped into the well is lost to the formation. If loss of circulation occurs in drilling or well servicing operations, the drilling fluid level may drop and, depending on the severity of the losses, the hydrostatic overbalance in the well will then be reduced or eliminated and primary control will be lost. The consequences of lost circulation vary from mild inconvenience to a serious Page 46 fH IWGF Training V 1.00threat to the safety of the rig. This mainly depends on the pressure regime in the well and the fluids contained in the formations which are exposed in the well bore. Losses are usually classified as follows: * static losses down hole losses which occur under static hole condition, e.g. no circula~ tion or string movement * dynamic losses down hole losses which occur during circulation. The difference between drilling fluid volume in/out will indicate the magnitude of the losses © total losses or lost cireulation no returns will be observed Causes of circulation losses Lost circulation is the loss of both liquid and solid components of the mud from the well bore to the formation . For this to occur, two main conditions must be met ; firstly the hydrostatic pressure of the mud must exceed that of the fluid in the formation and secondly the openings in the formation must be larger than the largest particles in the mud. A pressure difference sufficient to result in losses may be a result of pressure- depleted formations, about which nothing can be done, or by operational prac- tices such as high drilling fluid densities, excessive pipe running speeds, excessive high annular friction losses, balling up of bottom hole assemblies or excessive pressures caused by breaking circulation. There are a number of formation types that typically have openings large enough to allow lost circulation. Coarsely permeable unconsolidated formations Studies show that these formations only take mud if they have a very high permeability. Some upper hole sand and gravel beds have this character- istic and the fluids they contain are rarely abnormally pressured. In such formations, however, fluid may penetrate from the mud into the rock matrix leaving behind a wall cake which can plaster off the formation and prevent further losses. ‘* Cavernous or vugular zones ‘These zones are usually found in carbonaceous rocks such as limestones. The position and depth of such formations can often be predicted from geological maps. The size of the cavern or fissure may vary from the size of a pin up to the size of a large tunnel and usually it will be filled with water. When you penetrate such a zone the bit will often descend very rapidly and a sudden and severe loss of mud may be observed. * Natural fractures A natural fracture may be defined loosely as the interface between two WCF Training Page 47 v1.00layers of rock which have little or no chemical bond between them. When critical pressures are reached such fractures may open and take mud. Onee a fracture has started to take mud it can widen and take more mud at lower pressures. When the layers of rock are horizontal the over-burden forces them together and the likelihood of the fracture widening is reduced. Natural fractures are therefore more frequently encountered as being a problem in steeply dipping formations. * Induced fractures ‘The major cause of induced losses is mechanically imposed pressures which may result, for example, from rapidly running or spudding drill pipe, or rapidly reciprocating casing. Anything which closes the annulus, or reduces annular clearance, may also cause pressure build-up. For example sloughing shale may close off the annulus. In such cases pump pressure is then added to the effective hydrostatic pressure and formation breakdown may occur. You should note that if mud was lost to sloughing shales it can flow back into the hole once it is cleared. This will give the impression of a kick. There are several other factors which induce fractures. When excessive gel strengths are present the mud becomes difficult to move and the annulus behaves as if it is "closed off". Raising the mud density too rapidly or opening the pumps too rapidly can also cause pressure surges which may fracture the formation. Finally, if the hydrostatic pressure of mud in the annulus is larger than the formation pressure, fracturing may occur. Locating lost circulation zones Contrary to popular opinion, field investigations of such losses show that the majority of them are not found at the bottom of the hole. In fact, almost half of them are found at or near the bottom of the last string of casing. The way in which the losses started will often give an indication of their location and first attempts to cure them will often be based on an assumption of where they are occurring. However identifying the location of the losses more accu-, rately can greatly assist attempts to cure them if these first measures fail. Several methods are available for locating loss zones, including the spinner survey, the temperature survey, the radioactive survey and the pressure transducer survey. These all rely on detecting a sudden change in the proper- ties of the mud column (e.g. it's temperature or velocity) as a survey tool logs the well, indicating that the loss zone has been found. The choice of survey type is likely to be governed by what tools are available and the principles of each type will not be covered further here. One thing to note, however, is that they all require the drill string to be pulled out of hole before they can be run. As will be seen later in this part, it is much easier to control a well when the bit is as deep as possible and serious consideration must be given to the well control issues before pulling out of a well with losses. Page 48 INGF Training v1.00Restoring circulation There are a number of ways to go about reducing losses and / or restoring circulation. The choice of method will depend on the magnitude of the losses, the material and personnel available and the urgency i.e. a kick may be likely to occur if the losses continue or mud stocks at surface may be low. * “Pull up and wait” Pull up and wait is only used occasionally. It is applied in cases of seeping losses, i.e. 0-2 to 2 mh (1 to 10 bbV/h) or partial losses, i.e. 2 to 8 mh (10 to 50 bbW/h). Greater losses or total losses usually require more positive action. The "pull up and wait” procedure relies on the loss zone being sealed naturally by mud and cuttings. Circulation is stopped at the first indication of partial losses and the bit is pulled to a point of safety within the casing shoe. The bit should be pulled carefully in order to minimise surge or swab pressures which could further damage a thief zone. The hole is then left static for 4 to 8 hours and mud losses are made up by filling the annulus. Water may sometimes be used to fill the annulus (see below) but if s0, the fill should not extend below the casing shoe. After waiting, the bit is returned to bottom and attempt made to break circulation. If losses recur other procedures may be used. The facts that rig time tends to be more expensive than lost circulation material (LCM) and that the resulting cure is unlikely to be very strong make this method a poor choice when compared with using LCM immediately. © The use of lost circulation material (LCM) This technique is again used in the case of seeping or partial losses, either after or instead of the "pull up and wait" technique. After either making an assumption about the location of the loss zone or locating it, a quantity of LCM is mixed with 8 to 16 m$ (50 to 100 bbl) of mud. This pill is then displaced rapidly down hole and spotted opposite the loss zone. The drill string is immediately pulled back to avoid getting stuck should a bridge form and the LCM is left to soak for 4 to 8 hours. LCM is designed to plug the holes in the formation through which the losses are occurring and thus the type of LCM chosen will depend on the formations exposed in the open hole. If large fractures are thought to exist, it is some- times a good idea to pump coarse material to form a bridge and then follow this with a second pill of fine material to plug the holes in the bridging material. The concentration of LCM should be 60 to 90 kg/m® (20 to 30 ppb) if it is to be pumped through the bit nozzles. The same restriction will apply to the choice of medium or coarse materials as large calibre LCMs would tend to block the narrow bit nozzles. They are therefore only used when pumping through open drill pipe. There is a wide variety of materials that can be used as LCM, such as mica flakes, nut shells, graded salt and calcium carbonate. ‘The advantage of the last of the last two materials is that if they are used when drilling the reservoir section, they can subsequently be dissolved with water or acid respectively, thus avoiding impairment. Further information on LCM pills can be found in the WEDLP, Section 6, Part 1 - Drilling fluids. IWGF Training Page 49 v1.00© Squeeze cementing If the use of LCM alone has been unsuccessful cement squeezing tech- niques may be tried to seal off loss (or "thief") zones. All these techniques are similar, the main difference being in composition of the cement slurry. As with the use of LOM, the chosen slurry is mixed, pumped down hole and spotted opposite the loss zone. If the loss rate is low, the BOP is then closed and the slurry is squeezed into the zone by pumping slowly. The surface pressure will be limited to avoid inducing fractures if possible. ‘Thereafter the string is pulled back, circulation is broken to ensure that no cement remains inside the string and the cement is then left to set. It is often necessary to repeat the squeeze several times before the formation can be made to withstand high mud gradients. If the loss rate is high, as with large natural fractures or vugs, the main consideration will not be trying to push the slurry into the formation but trying to prevent it being flushed too far from the well bore to be effective. Assuming that previous LOM pills have failed to slow the losses, this can only be achieved by reducing the hydrostatic head of the annular fluids (cement, mud and water) to a minimum. Alternatively, it is possible to use a packer such as the Halliburton RTTS combined with a glass fibre stinger. The packer is set before cementing and the slurry spotted and allowed to set with the stinger in place. The packer prevents the full hydrostatic head of the mud column being transmitted to the loss zone. Once the cement has set, the stinger is broken and the plug is subse- quently drilled out. The coment slurries used for curing losses in this way are generally designed to have a short thickening time and a high water loss. In addition they may be mixed using thixotropic chemicals and / or LCM materials. Cement is the first choice for preparing a slurry as it is generally easily available at the well site and people are familiar with handling it. Alternative slurries are sometimes used when the problems are severe. Examples are high filtrate slurries or gunk slurries. The latter is mixed from diesel oil and bentonite, sometimes with the addition of cement. When this mixture comes into contact with water in the loss zone it rapidly forms a gel with a very high shear strength. The addition of cement means that this gel with also develop a high compressive strength. Great care must be taken when mixing and pumping gunk as if it contacts water in the surface lines they are easily plugged, compounding the problems. * Plugging Plugs are used to seal off high pressure zones when losses occur above them, possibly while killing a well that has kicked. Plugs are covered farther in Topic 6 of this Part, Tertiary Well Control. © Reducing the drilling fluid gradient Whenever losses occur, the first action should be to check whether the mud Page 50 IWGF Training v1.00gradient can be reduced. In principle, the lowest mud gradient which allows a sufficient trip margin should already be in use to maximise ROP and minimise the chance of differential sticking. However, it is possible that the formation which requires the programmed mud density to control it may not yet. have been penetrated. Alternatively, a high mud gradient may be in use in order to improve borehole stability. A reduction of mud density may facilitate subsequent operations with LCM or cement, or may allow drilling to continue to a suitable casing point so that a contingency casing string can be set. * Drilling with aerated fluid or foam In an area where it is known that no over pressured zones exist, a hole section could be drilled with aerated mud or foam. Aerating the mud reduces the effective mud density and allows drilling with returns to be resumed. This may be the only way to drill formations with very low pres- sures e.g. highly depleted reservoirs, as otherwise fracturing of differential sticking problems may be too severe. It is, however, a costly solution to the problem of lost circulation, not only because of the cost of aerating the mud, bus also because of potential corrosion problems. Inhibitors should be used to minimise corrosion and the pH of the mud kept at a minimum value of 12.5. Where a complete loss of circulation occurs, the level of fluid in the annulus will sometimes drop very quickly until the remaining hydrostatic head of mud in the annulus equalises with the pressure in the loss zone. This can create a potentially dangerous situation as other formations may then become under balanced and start to flow. In such situations it is important to try and maintain at least some pressure in the top part of the hole. This may be done by filling the annulus with water. If this is possible, and the well remains quiet, it may be possible to reduce the mud gradient sufficiently to be able to resume drilling with mud returns. When sudden and complete losses occur, drilling is immediately stopped and the kelly pulled back above the rotary table to facilitate closing the BOP if necessary. Simultaneously the annulus should be filled with water (or diesel for oil based drilling fluids). The volume added should be recorded. The mud gradient that the loss zone can support may then be calculated by working out the hydrostatic pressure provided by the fluids in the annulus. With luck, the rate of losses will slow and the well will become static. If the well remains quiet, the string can be pulled to the casing shoe while filling the hole with mud through the string. If the mud gradient that the loss zone supports is feasible, the well can now be circulated to that mud in stages. The supervisor needs to be very wary of the chances of getting stuck, since the reduction of pressure in the well bore may cause weak formations to cave. Also, an apparent reduction in losses at surface may merely be an indication that the well has started to flow. Drilling without mud returns If circulation cannot be established, and well control problems do not occur, then two methods of drilling ahead without returns may be considered: WCF Training Page 51 v1.00 @+ Blind drilling In this case, drilling continues at the minimum possible pump rate, to conserve mud stocks, while the annulus is topped up with water. If returns begin during drilling this might indicate that the loss zone has been plugged with drilled cuttings. However, it could also indicate the onset of a Kick. In some areas blind drilling is common practice as experience has shown that it is a waste of time to attempt to cure losses. Sometimes drilling is carried out with water, with occasional viscous pills to aid hole cleaning, without filling the annulus. It is even possible to do this when drilling frac- tured carbonate gas reservoirs by continually pumping water down the annulus at a flow rate sufficient to prevent gas percolation. The water supply is then safety eritical ! * Drilling with a floating mud cap ‘This can be used as a non-conventional alternative to blind drilling when an abnormally pressured formation occurs above a lost circulation zone in the same interval of open hole. Mud cap drilling controls the hydrostatic pressure on the loss zone by using a column of mud "floating" on a column of water. The mud column is used to maintain primary control over the abnormal pressure zone. Water is pumped down the drill pipe to cool the bit and clean the hole. Simultaneously the annulus is kept full with mud. A delicate pressure balance is maintained and, if managed correctly, it is possible to keep the annulus almost static by varying the flow rate of water being pumped down the string. Prior to any trip the drill pipe is displaced with mud. An NRV or Drop In Sub (DIS) should be installed in the string when drilling with total losses or with a floating drilling fluid cap. The dart of the DIS should be dropped and pumped down/latched prior to each trip. 1.4.3.2. PRIMARY CONTROL WHILST TRIPPING 1.4.3.2.1. Swabbing As mentioned under 1.4.2.1 swabbing is a reduction in bottom hole pressure (BHP) caused by moving the drill string upwards. It can be categorised into either low volume or high volume swabbing. Low volume swabbing is rela- tively easy to prevent and does not usually create difficulties. High volume swabbing, on the other hand, is one of the prime causes - if not the major cause - of blowouts. Whilst tripping out of the hole high volume swabbing is the result of the BHA acting as a piston within the bore hole. It can cause a reduction in the bottom-hole hydrostatic pressure both directly or due to the flow of drilling fluid out of the drill string into the low pressure zone below the offending component. The annular restriction will prevent the annular hydrostatic pressure above the component from being transmitted past it and so the Page 52 INCF Training v1.00bottom hole hydrostatic pressure will be determined by the height of the remaining drilling fluid in the drill string. In large diameter holes, the ratio of open hole to drill string capacity is very high resulting in a rapid reduction in the fluid height within the drill string and consequently a rapid reduction in bottom hole fluid pressure which in tan may allow an influx from any porous and permeable formation below the restriction. Swabbing while tripping is doubly hazardous in as much as there is not only an influx into the well, but the influx is trapped below the BHA. If the influx is gas it does not even have to be very large to create a major hazard as it will only start to expand significantly as it approaches surface. This means that if the crew is not alert they will only become aware of the problem as the BHA reaches surface, by which time it is far too late to be able easily to regain control of the well. Circulating out a kick with the DP on bottom is never to be underestimated, but it is a routine operation - regaining control of a well when the drill string is almost out of the hole is a major exercise with a high probability of failure. Identification Safety requires continuous monitoring of topping up and return volumes throughout every round trip. This should be done with a frip tank and trip tank chart recorder or trip sheet. A short round trip or eheck-trip followed by circulating bottoms up before pulling out of the hole completely will provide the most reliable information on the swabbing characteristics of the system. This might indicate whether pulling speeds anW/or drilling fluid properties should be adjusted. When partial losses are experienced and a round trip is being made, the degree of losses should be accounted for in the trip tank measurements during the round trip, otherwise swabbing may not be detected. Prevention Three principal factors contribute to a swabbing tendeney : * annular clearance + pulling speed * drilling fluid properties Annular clearance is determined by the drill string and assembly configu- ration for a given hole size. However, hole size can be reduced by excessive mud cake, swelling formations, or an accumulation of cuttings (i.c. on low side in deviated holes). With a large annular clearance, the tendency for swabbing to occur is reduced. The clearance can be dramatically affected by balling of cuttings around the bit and stabilisers in the string. Low pipe speeds reduce the possibility of swabbing by reducing the pressure drops across the pipe length which are proportional to the velocity of the fluid relative to the pipe. IWGF Training : 7 Page 53 v1.00 @Overbalance is the main factor in reducing the possibility of swabbing. Controlling the yield point of the drilling fluid is also very important, since the swabbing effect is almost proportional to it. Minimising the solids content will help maintain a low yield point. A low water loss assists in obtaining a thin filter cake on the borehole wall, and in minimising the swelling of shales. This will reduce the balling-up of bit and stabilisers and thus reduce swabbing tendencies. When swabbing is anticipated it is recommended not to pump a heavy pill prior to commencing a round trip. This permits a more accurate check for swabbing. The heavy pill can be pumped when it is certain that swabbing will not occur. If low pipe pulling speeds and good drilling fluid properties are not sufficient to allow a round trip to be made safely, consideration should be given to circulating whilst the string is pulled, or to increasing the overbal- ance. It is important to pump at a rate to prevent any loss of overbalance, particularly in large diameter holes. 1.4,3.2.2 Failure to Fill the Hole Properly When the drill string is pulled out of the hole it is necessary to replace the steel volume removed from the well with drilling fluid if the fluid level in the well (and thus bottom hole pressure) is to be maintained. The volume required is equivalent to the open end displacement volume of the tubular being removed from the well. This volume, expressed usually in bbI/ft or m°/m, is provided in Drilling Data books. It is important to use the correct grade of tubular (i.e, wall thickness) when determining the displace- ment volume. Whilst tripping it is standard practise to keep the hole fall by circulating across the bell nipple. A small pump transfers fluid from a small, calibrated tank, the trip tank, and the flow from the well is taken back into the tank. As pipe is pulled from the well the volume of fluid in the tank reduces and accurate measurement of this change is compared to the theoretical volume of steel removed from the well. If the reduction in the tank is less than the theoretical volume then the well is not taking the correct amount of fluid and swabbing or an influx may be occurring. Ifthe change exceeds the theoretical volume, then the well is taking too much and there are losses occurring. 1.4.3.2.3 Tripping Practices Trip tank and trip sheets The hole should be kept fall at all times. Consequently, when tripping the drill string or running the casing both in and out of the well, the trip tank should always be used for the accurate measurement of drilling fluid volume changes. Page 54 IWCF Training v1.00Pulling liquid-filled pipe, a wet trip, should be avoided if possible. If it is unavoidable a “mud box” should be used to catch the drilling fluid from the drill pipe and return it to the well or the trip tank. (picture). It is essential that the volume changes in the trip tank equal the theoretical changes in the steel volume being added or removed from the well. This is the responsibility of the driller "on the brake" and must not be delegated to a subordinate. Whilst tripping, a record, called a trip sheet, should be kept of trip tank volumes against pipe displacements of each stand, to identify any discrepan- cies (e.g. gains or losses) as soon as possible. Ideally, the trip sheet should be prepared prior to the trip and has the theoretical volume per stand or 5 stands marked on it. The driller then only has to add the actual trip tank measurements to get a simple check of the hole behaviour. Flow checks Flow checks should be made as a minimum, before pulling out, in the last casing shoe, and before starting to pull the drill collars. For exploration wells it is recommended to make a short check trip prior to pulling out completely to verify that the overbalance is sufficient to prevent swabbing the well. After the check trip - of perhaps ten stands - the well is circulated "bottoms-up" to check for gas cut drilling fluid or other signs of near balance. Running speeds Pipe pulling or running speeds should be controlled to limit pressure surges. which otherwise could cause a kick, losses, or formation damage. Pressure surges are evaluated by considering the hole dimensions, string configuration, drilling fluid characteristics and pipe velocity. For normal hole conditions and pulling speeds, surge pressures are of a similar order of magnitude to annular friction losses whilst circulating. Critical sections In order to avoid swabbing effects, the drill string can be circulated whilst pulling through critical open hole sections. This is one of the fundamental advantages of the top-drive system over the conventional Kelly/Rotary Table system. Circulation can be started at any time and the string can still be pulled in stands rather than in singles that have to be laid down and then. picked back up again later. Swabbing In circumstances where it is suspected that swabbing may have occurred, the pipe should be run back to bottom, bottoms circulated up and the drilling fluid conditioned if necessary. If the well is flowing after swabbing, and the influx is above the bit, it should be closed in and killed at the prevailing bit depth. If the influx is below the bit the pipe should be stripped back to bottom IWGF Training : ___ Page 55 1,00 @using the combined volumetric stripping process. It must not be attempted to run back to bottom with an open flowing well. Time out of hole The time out of the hole should be kept to the minimum, consistent with oper- ational requirements and with regard to the hole stability. The hole should be kept full and monitored continuously. Similarly, if rig equipment has to be repaired, the pipe should be run or pulled to the shoe of the last casing string before repairs begin. In well servicing operations, the pipe should be run to sufficient depth to ensure that the well can be controlled. Circulation at intermediate stages Except for specific reasons (e.g, when retrieving long and tight formation cores), circulation should preferably not be carried out at intermediate stages when tripping drill pipe or running casing into the hole. A flow check is preferred. Records Accurate records of all operations must be maintained, particularly of any unusual occurrences. These should describe alll flow line shows. The duration of the occurrence, drilling fluid return gradients, chloride content, ete. should be also noted. 143.3 SHALLOW GAS Further reference may be made to the SIPM Shallow gas procedures guidance manual (EP 88-1000). Since, by definition, it is not possible to close in a shallow gas flow and imple- ment true secondary pressure control measures it is imperative that primary control is maintained. Common drilling practices which are applicable for top hole drilling in general and diverter drilling in particular are summarised below. Recommendations are made with a view to simplify operations, thereby minimising possible hole problems. * Apilot hole should be drilled in areas with possible shallow gas, because the small hole size will facilitate a dynamic well killing operation. * The penetration rate should be restricted to avoid an excessive build-up of solids in the hole which could cause formation breakdown and losses. Drilling with denser drilling fluid returns could also obscure indications of drilling through higher pressured formations. In these circumstances the well could kick while circulating the hole clean. Restricted drilling rates also minimise the penetration into the gas bearing formation which in turn minimises the influx rate. An excessive drilling rate Page 56 IWCF Training V 1.00through a formation containing gas reduces the hydrostatic head of the drilling fluid, which may eventually result in a flowing well. + Every effort should be made to minimise the possibility of swabbing. Pumping out of the hole at optimum circulating rates is recommended for all upward pipe movements (e.g. making connections and tripping). Especially in larger hole sizes (i.e. larger than 12") it is important to check that the circulation rate is sufficiently high and the pulling speed sufficiently low to ensure that no swabbing will take place. A Top Drive System or Cireulating Drill Pipe Elevator (based on the Regan Fast Shut-off Coupling) will facilitate efficient pumping out of hole operations. ‘The use of stabilisers also increases the risk of swabbing, hence the minimum required number of stabilisers should be used. * Accurate measurement and control of drilling fluid is most important in order to detect gas as early as possible. Properly calibrated and func- tioning gas detection equipment and a differential flow meter are essen- tial in top hole drilling. Flow checks should be made - before tripping - at any time a sharp penetration rate increase or tank level anomaly is observed (if returns to surface are taken) - when any anomaly appears on the MWD log (if an MWD data trans- mission system is used) - at any specific depths referred to in the drilling programme (taken from the shallow seismic survey). It is recommended to flow check each connection while drilling the pilot hole in potential shallow gas areas. Measuring drilling fluid density in and out, and checking for seepage losses are all important practices which must be applied continuously. * A float valve (NRV) must be installed in all bottom hole assemblies (BHAs) which are used in top hole drilling in order to prevent uncontrol- lable flow up the drill string. The float valve is the only down-hole mechanical barrier available. The use of two float valves in the BHA may be considered in potential shallow gas areas. * Large bit nozzles or no nozzles and large mud pump liners should be used to allow lost circulation material (LCM) to be pumped through the bit in case of losses. Large nozzles are also advantageous during dynamic killing operations, since a higher pump rate can be achieved. As an example, a pump rate of approximately 2700 V/min at 20,000 kPa pump pressure can be obtained using a 1,300-1,600 HP pump with 3 x 14/99" nozzles installed in the bit. By using 8 x 1832" nozzles, the pump rate can be increased to around 3,800 l/min at 20,000 kPa, (Values will vary slightly depending on drilling fluid density, hole size, BHA configuration, ete.). The use of a centre nozzle bit will increase the maximum circulation rate IWF Training = ee eet Page 57 v1.00still further and also reduces the chance of bit balling. * Shallow kick-offs should be avoided in areas with probable shallow gas. Top hole drilling operations in these areas should be simple and quick to minimise possible hole problems. BHAs used for kick-off operations also have flow restrictions which will reduce the maximum possible flow through the drill string considerably. A successful dynamic well killing operation will then become very unlikely. Page 58 IWCF Training v1.00Topic 1.5 Secondary well conirol 1.5.1. PRINCIPLES Secondary control is the proper use of blow-out prevention (BOP) or pressure control equipment to regain control of the well in the event that primary control cannot be properly maintained. The configuration of such pressure control equipment is covered in Part 2 - “Well control equipment” and Part 3 - “Sub-sea well control vies equipment”, surface pressure The pressure control equipment Y prevents flow from the well and allows surface pressure to develop in an underbalanced situation. The surface casing pressure will rise to the point where it equals the bottom hole pressure less the hydrostatic head of the fluid in the annulus (where a string of tubulars are in the well) or the hole (where there are no tubulars in the well) - see Figure 1.14. Where a drill string is in the well and pressure Driling control equipment has been hid gracion? installed on it, the drill string surface pressure will rise until it equals the bottom hole pressure — less the hydrostatic head of the Underbaiance at / fluid in the drill string. reservoir depth Figure 1.14: Pressure balance with BOP equipment closed. Pressure Reservoir gradient Depth Well control equipment permits the well to be closed in and remedial action to be taken, usually in the form of removing any influx of gas, oil or water, and restoring primary control. Where the influx has been caused by the penetration of an overpressured formation this will entail the controlled displacement of the well to drilling fluid with an increased gradient. In the case of swabbing where the kick has been caused by the swabbed-in influx reducing the hydrostatic pressure, it is only necessary to remove the influx under controlled conditions to regain primary control of the well. IWCF Training V 1.00From the point that the well is closed in until primary control has been re- established, the first objective of all operations carried out is to maintain bottom hole pressure at, or only slightly above, the formation pore pressure, Po, Allowing it to drop below Po will allow a second influx. If the bottom hole pressure is any higher than Py then there is an unnecessary risk of causing damage to the formation. This is one of the basic tenets of well control. Circulation whilst holding dynamic back pressure via a variable choke is the most common method for removing an influx in a controlled manner. The amount of back pressure, measured by a pressure gauge immediately upstream of the choke (and called the choke pressure or Pz,) is adjusted such that the bottom hole pressure is maintained at a constant value equal to or just above the formation pore pressure. Due to the presence of the bit/nozzles it is rare for an influx to enter the drill string and so this conduit to the bottom of the hole, containing a fluid of a known gradient, provides a simple way of monitoring both the dynamic and static bottom hole pressures whilst killing the well in a controlled manner. 1.5.1.1 RECOGNISING A KICK A kick occurs when the hydrostatic pressure of the drilling fluid column in the well is less than the formation pore pressure (i.e. there is an underbal- ance), provided that the formation has the ability to produce. Whereas a drilling break and gas, salt water, or oil contaminated drilling fluid at the flow line are cautionary signals, a kick provides definite information that formation fluid is entering the well bore. When a well kicks, an immediate increase in drilling fluid returns and volume in the active system will be observed. The size and severity of the kick depend upon: * the degree of underbalance; * the formation permeability; * the length of time the formation is allowed to remain underbalanced 2 NEED FOR IMMEDIATE ACTION When a well kicks, it should be closed in immediately. Large influx volumes and correspondingly large fluid volume increases at the surface may be obtained in a very short time. The larger the influx, the higher will be the annular pressures, both initially and during subsequent well control opera- tions. It is therefore imperative that a kick is contained as early as possible. Since the reaction time is critical, equipment should be suitable and crews should be trained to detect small changes in the active drilling fluid volume immediately the changes occur. The proper method of closing the well should be rehearsed so often that it can be done quickly and without mistakes. This exercise is called the well control drill. Increased drilling fluid returns are not always conclusive evidence that the Page 60 WCF Training v1.00well is kicking. However, there should always be a flow check and if necessary ‘THERE SHOULD BE NO HESITATION IN CLOSING THE WELL IN. The decision to take immediate corrective action is the responsibility of the driller. 1.5.1.3 THE “BALLOONING” EFFECT The plastic behaviour of shales in the overbalanced mode, so-called "borehole ballooning" can cause similar symptoms to those of a kick in overpressured formations. Major manifestations of ballooning are: * initial high gas shows from recently drilled sand layers, regardless of their much lower original natural pressures; initial loss of drilling fluid when circulation is started, followed by an equivalent gain of fluid after circulation is stopped; CIDPP equal to the ECD effect on BHP, immediately after circulation is stopped and the well is closed-in; «high penetration rates with very low WOB. . . Genuine symptoms of underbalance may also be caused by nearby wells which have been drilled with a large overbalance. Side-tracked holes often behave quite differently from the original hole for the same reason. Reducing or stopping the ballooning effect is probably best achieved by bleeding off pre-existing charged shales as drilling progresses, sticking close to the natural pore pressures of the formation. It will result in temporary shows and brittle sloughing. "Borehole ballooning" is explained in the Oil & Gas Journal, 13 March 1989, 1.5.2 PLANNING A loss of primary pressure control that requires the implementation of secondary pressure control can be a stressful situation. Whilst the procedures for restoring primary control are straightforward they involve the use of several sets of data, much of which can be prepared in advance; reducing the risk of errors made under stress. 1.5.2.1 KICK CONTROL WORKSHEET A Kick Control Worksheet is used to pro-record data relating to the well required for secondary control. The sheet allows the calculation of all volu- metric data needed during the killing of a well. It provides an easy to follow method of reacting to a kick, taking the user through the calculations required after a kick to plan and then implement the recovery operation. An example of this sheet can be found in Appendix 8 of EP 89-1500, the Pressure Control Manual for Drilling and Workover Operations. During drilling operations, the Kick Control Worksheet should be updated regularly to correct depth-dependent data and additionally whenever signifi- IWCF Training Page 61 v1.00cant changes occur in parameters such as casing and hole sizes, drilling fluid gradient, drill string configuration, nozzle sizes, formation characteristics, ete. The frequency of revisions should be determined to suit local conditions and the extent of knowledge regarding pore pressures. For exploration or appraisal wells, kick control worksheets should be available at all times that BOP are installed and in use. They should be updated at least daily or every 150 m (500 ft), whichever is the sooner. 1.5.2.2. MAASP DETERMINATION During well control operations the maximum allowable borehole fluid pressure at the casing shoe is normally considered to be the critical factor, based on the assumption that the weakest formation is at the shoe. For prac- tical purposes, the corresponding pressure at the surface is calculated and used during operations since it can be measured directly. This value, the Maximum Allowable Annular Surface Pressure (MAASP), was discussed in Topic 1.3.6. Data from leak-off tests is combined with the current drilling fluid gradient to generate the MAASP value. Some control panels of remotely controlled chokes have an automatic MAASP control feature which doos not allow the choke pressure Pa, to exceed the pre- set MAASP. When this feature is used in situations whereby MAASP should be exceeded, for instance when top influx is already inside the casing shoe, more influx is allowed to enter the hole unnecessarily which may lead to worsening situations. It is recommended not to use this feature, but to adjust the remotely controlled choke(s) manually from the control panel. 1.5.23 KICK TOLERANCE The kick tolerance is the maximum size of influx that can be safely circu- lated out without breaking down the formation just below the previous casing shoe (see also the discussion of maximum safe drilling depth in Topic 1.3.5). An assumption has to be made about the kick tolerance during the design of a well, based on the following variables: the hole size geological control, * any known PVT characteristics of reservoir fluids. © the accuracy and reliability of the available kick detection system © the type of rig (floater/bottom supported rigs) © the experience of the crew Likely influx volumes for the drilling and reservoir conditions prevailing can be calculated/determined from equations given in the Casing Design Guide. However, results largely depend on realistic values of kick detection threshold volumes for the rig used to drill the well. Experience indicates that well- trained crews can perform BOP drills in three to four minutes. BOP drills should be carried out to achieve and maintain this performance standard. Page 62 IWCF Training v1.00Whilst drilling the well periodic checks should be made to establish the magnitude of the influx which can still be safely handled with the drilling fluid currently in the hole without causing formation failure (i.e. kick toler- ance). Two design cases may be considered: * maximum swab kick (assuming gas) that can be taken using the existing drilling fluid density * maximum gas influx volume assuming different degrees of overpressure relative to the current drilling fluid density. These checks will indicate the extent to which the density can be increased to control formation pressure or whether casing should be set prior to drilling deeper. New calculations should be made whenever any of the parameters have, or will be, changed. 1.5.2.4 SLOW CIRCULATING RATES ‘Well control operations are performed at reduced pump speeds in order to: * allow weighting up and degassing of the drilling fluid * reduce loads on surface equipment * increase reaction time for correct choke adjustments. Operation of the pump at a pressure too close to the setting of its relief valve is dangerous. If the relief valve opens, drilling fluid will flow back through the drill pipe and standpipe manifold. Circulation pressures at the reduced pump speeds must be known in advance of a well killing operation. Before starting to drill with a new bit and at the beginning of each shift, the driller should therefore observe pump pressures at selected slow circulation rates and record the results in the daily drilling report and kick control worksheet. Circulation pressures may differ considerably when using a different equip- ment hook-up (eg. circulating head/chicksans vs. kelly or top drive). Slow circulation rates should be taken with the equipment hook-up which is planned, or most likely to be used during the well killing operation. Slow circulation rates and pressures can be plotted as a straight line on a graph using log-log paper. Appropriate pump pressures for any pump rate can then be estimated using this graph, provided the drilling fluid properties and drill string configuration do not change. When using different drilling fluid properties, the slow circulation rates should be repeated. If necessary, provided the rheology has not changed too greatly, pressure losses may be approximated using the original figures as. circulating pressure with original drilling fluid circulating pressure with new drilling fluid original drilling fluid gradient new drilling fluid gradient IWGF Training f Page 63 v1.00 @15.2.5 RIG MIXING CAPABILITIES When killing a well using the balanced drilling fluid or “Wait & Weight” method, it is necessary to increase the density of the drilling fluid returning into the active system from the degasser prior to it being pumped into the well. Thus the maximum circulating rate can be determined by the rate at which weighting agent, normally barytes, can be added to the system. The weighting up capacity depends on the maximum rate at which barytes can be added over prolonged periods. It is recommended that this rate is established for drilling fluid gradients in normal use. An important advantage of the balanced drilling fluid method is that a well can be controlled in one circulation. This is done by employing a constant pump speed whilst circulating a drilling fluid which has been weighted up. If fluid mixing facilities cannot achieve the required increase in gradient, it will not be possible to control the well in one circulation. Several circulations may be required, the gradient being raised in stages during each circulation, or on a gradual continuous basis while pumping. One option to obtain the required drilling fluid density quickly is to prepare Kill fluid prior to drilling ahead in possibly overpressured formations. This practice will facilitate speedy well killing operations. Preparing kill fluid in advance is recommended for wells in which increased densities are expected to be used. The recommended kill fluid gradient depends on the expected formation pressures to be encountered, but as a general rule may be 10-20% higher than the actual drilling fluid density in use. The correct density can be obtained quickly by diluting the kill fluid with the original drilling fluid 1.5.2.6 CONTINGENCY STOCK LEVELS Minimum stock requirements for weighting materials, chemicals, lost circula- tion materials, cement, drilling fluid and water (supply) on the well site depend on the well prognosis, severity of potential drilling problems, rig load capacity (offshore), remoteness and accessibility of the well site, and avail- ability of materials. Proper stocks to allow well control operations to be carried out should be provided in all cases. Particular attention to this aspect should be given on remote onshore and all offshore locations. The following should be included: * adequate liquid storage; © adequate dry drilling fluid chemicals, barytes or weighting material and drill water storage; * sufficient on-site cement storage (sufficient for at least 3 x 150 m open hole cement plugs). ‘When there is no water directly available, water may be trucked or (a) water well(s) may be drilled. In both cases, ample water storage should be available with a minimum recommended water delivery to the rig site of at least 30 m® ‘nour. The minimum required water storage capacity should be based on the Page 64 IWGF Training v1.00reliability of a continuous water supply. A 500m? spare water pit is suffi- cient to allow for a 24-hour supply interruption. Properties of reserve drilling fluid should be checked and maintained daily. The inventory of materials on site should be reviewed daily and replenish- ment arranged immediately when stock levels approach the specified minimum requirement. 1.5.3 REGAINING CONTROL 1.5.3.1 SHALLOW GAS By definition it is not possible to close in a shallow gas kick as the formation strength of the previous casing / conductor shoe does not provide sufficient strength. Consequently there remain two actions that can be taken, generally in tandem: * Diversion of the flow either at surface or sea bed as appropriate to protect personnel and equipment from the flow of gas © Attempt to dynamically kill the well 1.5.3.1.1 Diversion ‘The flow of water and gas is normally diverted either at surface or sea bed using an annular preventer mounted on a spool with one or more side outlets. (Refer to Part 2 and EP 88-1000, the SIPM Shallow Gas Procedures Guidance Manual) . It is essential to realise that these assemblies are not yet designed to with- stand the highly erosive flow of shallow gas for prolonged periods of time. The diverter assembly merely buys time to attempt a dynamic kill and/or for abandonment of the installation. 1.5.3.1.2. Dynamic kill A dynamic hill is a technique for controlling a blowout utilising the flowing frictional pressure to supplement the hydrostatic pressure of the kill fluid being pumped into the blowing well. The flow rate must be high enough and maintained such that the sum of the frictional and hydrostatic pressure exceeds the static formation pressure and the well ceases to produce. Critical features of a shallow gas kick are the speed of events linked to a rapid inerease in flow rate as the drilling fluid in the well is expelled by the expanding gas column. The earlier that a dynamic kill attempt can be initi- ated the greater the chances of success. The probability of success of the kill attempt is governed by: * the residual hydrostatic pressure of whatever fluid is in the well * the amount of back pressure that can be generated The remaining hydrostatic pressure is determined by the size of the influx IWGF Training 7 Page 65 Vv 1.00when killing starts. Due to the massive permeability and porosity of the shallow formations the flow rates, even at low draw down pressures, can be very large. Consequently, it is not unusual for the well to unload entirely before any reaction can be taken. This situation is exacerbated by the diffi- culty in identifying that the well is starting to flow, particularly when drilling riserless. ‘The amount of dynamic back pressure that can be generated is determined by: * hole geometry (annular clearance between drill string/BHA and the hole) * flow-rate of the kill fluid * kill fluid density and rheological properties 1.5.3.2 CLOSING IN THE WELL 1.5.3.2.1 Initial actions When a well kicks, it should be closed in immediately. Large influx volumes and correspondingly large drilling fluid volume increases at the surface ‘The decision to close in may be obtained in a very short time. The larger | the well is the responsi- the influx, the higher will be the annular pres- _| bility of the driller. This sures, both initially and during subsequent well _| should be initiated control operations. It is therefore imperative immediately and without that a kick is contained as early as possible. recourse to a supervisor ! Immediately an increase of flow from the flow line, or an unaccounted for increase in the level in the active tanks, is noticed the ongoing operations should be discontinued, the kelly (if drilling) raised until the first drill pipe tool ————_—— joint is above the rotary table Whenever the drill string is being pulled | (for platform and land drilling) or run in hole there should be suitable | and the pump stopped. A flow crossovers available on the drill floor to. | check should be made and, if allow a kelly-cock to be installed on any | there is flow, the well should be component in the drill string with the closed in - the different methods minimum of delay. Ideally, the crossover | that can be used to do this are described in Topic 1.5.8.2.3 should already be made up to the kelly- cock. The key for the kelly cock, which is | below. If the kelly is not used to close it, must also be available. installed then a kelly cock or The key is frequently modified by the rig_| Inside BOP (IBOP) should be crew in order to make it easier tohandle | installed on the drill string in a hurry. (with a suitable cross-over if required) prior to closing in the annulus. Having closed in the well the kelly or a suitable circulating head should be Page 66 IWCF Training v1.00installed (if not drilling) and the string positioned such that no tool joint is opposite the pipe rams. One set of pipe rams should then be closed. On most BOP stack configurations this is the set of pipe rams above the blind / shear rams. Although the annular preventer will often have a pressure rating well in excess of the closed in annulus pressure; this may change rapidly in the case of circulating out a gas influx. Of all the BOP’s installed the annular preventer is the weakest and most often gives problems. Therefore, closing a pipe ram at this early stage prevents the necessity of doing it later when pressures are much higher. 1.5.3.2.2 Drilling with oil based fluid (OBM) In principle, liquid hydrocarbons (such as diesel oil or base oil) can dissolve large quantities of hydrocarbon gases (methane), COs, and HS. If conditions (temperature, formation pressure) are such that gas has been dissolved in OBM and the influx is being circulated upwards, gas will begin to separate out when the bubble point of this mixture is reached. This means that a gas kick in OBM can initially behave like a water kick with spontaneous expan- sion of the gas close to surface when the dissolved gas comes out of solution. The most important aspect, when drilling with OBM, is to be aware of the unpredictable behaviour of gas expansion near surface, especially when circu- lating bottoms up after a round-trip or any circulation after long static periods, because gas may have entered the well before circulation started. Although the solubility of gas in OBM makes drilling with OBM more hazardous with regard to a possible sudden gas breakout near surface during normal circulation, well control with OBM is not jeopardised, because the principle of kick detection (pit gain, increased flow during the time of taking the kick, ete.) remains the same for all drilling fluid types. However, increased flow due to the expansion effect of gas rising in water based fluid is not seen in OBM in the early stages of taking the kick and small gas quantities (tight formation) will probably only be detected when the influx is close to surface. ‘This means that gas breakout and sudden unloading is a definite possibility. Drilling crews have to be made aware of this problem. Standard well killing procedures apply for OBM. However, gas present in gas cut OBM may be difficult to remove completely due to the solubility of the gas. If the presence of dissolved gas in OBM is suspected and it is expected to cause splash problems at surface, the well may be closed-in prior to reaching bottoms-up, after which circulation can be completed via the choke manifold. Wells drilled with a sub sea BOP stack need to be closed-in before the suspected influx has reached the BOPs. 1.5.3.2.3 Close-In methods A number of different close-in techniques are used within the drilling industry. Much confusion is caused by different names being applied to different (or the same!) methods. Discussions into the benefits and limita- IWGF Training Page 67 v1.00 @tions of the various methods persist. The preferred method of closing in wells within Shell Operating Units is the Hard Close-In method. The following methods are as defined in API RP 59 (Recommended Practices for Well Control Operations). Hard close-In For a hard close-in the choke remains closed at all times other than during a well control operation. The choke line valves are aligned such that a flow path is open to the choking system with the exception of the choke(s) itself and one choke line valve located near the blow-out preventer stack. When the hard close-in procedure is selected for closing in the well the blow-out preventer is closed. If the casing pressure cannot be measured at the wellhead the choke line valve is opened with the choke or adjacent high pressure valve remaining closed so that pressure can be measured at the choke manifold. Soft close-in For a soft close in a choke is left open at all times other than during a well control operation. The choke line valves are aligned such that a flow path is open through the choking system with the exception of one choke line valve located near the blow-out preventer stack. When the soft close-in procedure is selected for closing in the well © the choke line valve is opened ‘+ the blowout preventer is closed © the choke is closed Fast close-in Fast close-in is a sub-set of the hard close-in and is sometimes used to differ- entiate between the use of the annular preventer (fast) and a pipe ram (hard) to initially close in the well. This is by no means a universal definition and both “hard” and “fast” can be used to describe the use of either !! Hard (/fast) vs. soft close-In Posit API takes a noncommittal position in API RP 59. It advises that the hard close-in procedure allows the well to be closed-in in the shortest possible time thereby minimising the amount of additional influx of kicking fluid to enter the well bore. It further advises that the use of the hard close-in procedure be limited to well conditions in which the maximum allowable casing pressure is greater than the anticipated initial close-in pressure and a well fracture would not be expected to broach to surface on initial closure. The API states that the soft close-in procedure allows the choke to be closed in such a manner to permit sensitive control and monitoring of casing Page 68 IWCF Training v1.00pressure build-up during closure. This is especially important if formation fracturing is likely to occur if the well is closed in without regard to the possibility of excessive initial closed-in casing pressure. Shell Position: In almost all Shell Operating Companies the hard shut-in has been adopted as the preferred method of closing in a well. Reasons for this are * The influx volume and height is smaller than when a soft shut in (i.e. closing in the well on the choke after the BOP is closed) is used. * The pressure at the casing shoe due to the “water hammer” effect during a hard shut-in is less than that due to circulating out the larger kick which would be taken were a soft shut in to be used (KSEPL research). 1,5.3.2.4 Close-in procedures ‘To summarise the discussion above, if the well is to be closed in using the hard close-in method following a kick while drilling, the procedure is as follow. © Stop rotary. * Raise the kelly or top drive until the first tool joint is above the rotary table (this ensures that a tool joint is not opposite the pipe rams). Stop circulation, * Close the BOP If the kick occurs while tripping, the procedure is more complicated due to the necessity to install an internal blowout preventer on the drill string. The procedure is the same whether tripping into or out of the hole. Lower string so that the top of the stand is at the rotary table. © Set the string in the slips. Install a full opening ball valve, commonly referred to as a kelly cock, and close it. * Close the BOP. 1.5.3.2.5 Drills As discussed above, it is important that the close-in procedures are executed in the shortest possible time. It is therefore essential that the crew have regular practice in such procedures. Such exercises are called kick drills and trip drills. A Kick or pit drill is aimed at ensuring the alertness of the driller and is generally initiated by that persons supervisor. The latter will may give a verbal instruction to initiate the drill. Alternatively the supervisor may manually move either the return flow meter (Flo-Sho) or one of the pit volume totaliser transducers to simulate indications of a kick on the driller’s instru- mentation. The reaction time of the driller will then be measured. The objective of the trip drill is to reduce the reaction time of the crew toa IWF Training Page 69 v1.00kick while tripping. The drill is initiated verbally by the driller and the crew must follow the procedure above. 1.5.3.2.6 Monitoring surface pressures and volumes After closing in the well and lining the annulus up to the choke manifold, the pressure should be monitored on both the annulus and DP gauges until it stabilises (first build up). The time required for the pressures to stabilise will depend on the permeability of the producing formation. The stabilised pres- sures are required for subsequent calculations prior to killing the well. If the influx is a gas, or contains gas, this gas will migrate up the well. With the well closed in the gas can not expand and will retain approximately its original pressure. The column of fluid above the gas will become smaller as the gas rises and therefore the hydrostatic pressure due this fluid column will reduce. This will result in an increase in surface pressure on the annulus. Similarly, the column of fluid below the gas will increase with a resultant increase in bottom hole pressure and therefore also the drill pipe pressure (Pa,). A crew member must continue to observe the gauges to check for farther pressure increases as a result of the influx migrating up the annulus (second build up). As pointed out in Topic 1.5.1, the aim throughout is to keep the bottom hole pressure equal to Po. Thus, if the pressure increase due to a second build up becomes significant, fluid must be bled off from the annulus to allow the gas to expand as it migrates and thus keep the bottom hole pressure constant. An accurate measurement of the volume gained in the tank(s) must also be made at this point as this gives the volume of the influx in the well. This value is also necessary for subsequent calculations. 1.5.3.3 SITUATION ASSESSMENT Once the well has been closed in and control regained, there is time to assess the nature and severity of the influx. 1.5.3.3.1 Position of influx Before commencing any calculations it is important to have a clear idea of what the physical situation is in the well and, in particular, what the position of the influx is. It is strongly recommended that the first step in response planning is to make a well sketch if this does not already form part of the kiek control worksheet. If the kick occurred while drilling, the situation is generally considered to be as depicted in Figure 1.15 and this forms the starting point for most standard calculations. It should be realised, however, that in reality the influx will be more spread out since it will have begun while the drilling fluid was still in circulation. Furthermore, as stated above, a gas influx will tend to migrate upwards and, though generally considered as a single bubble for calculation purposes, will actually break up into a stream of bubbles or slugs. Page 70 IWF Training v1.00If the kick occurred while tripping then the situa- tion is more difficult to assess. Whereas a kick during drilling may reasonably be assumed to have come from the bottom of the hole, a swab kick could originate from any point in the open hole, probably but not necessarily from below the bit. If the formation fluid is gas, there may be two influxes. The first smaller influx enters the well when the well bore pressure is temporarily reduced as the drill string is momentarily pulled too fast or the bottom hole assembly is pulled through a tight spot. This influx then begins to migrate upwards and, since the well is still open, will expand as the pressure of the drilling fluid reduces. As it expands it occupies a greater height in the well displacing drilling fluid from the hole. As a result the hydrostatic pressure at the depth of the producing formation reduces at an increasing rate until the pressure in the well bore falls below Py and the formation starts to produce causing a second influx. A careful exami- nation of the trip sheet together with the prog- nosed open hole pressure profile will be required in order to estimate what the situation is. 1.5.3.3.2 Establishing formation pressure Figure 1.15 : influx while drilling Since the drill string is filled with drilling fluid of a known density (see Topic 1.5.1 ), the formation pressure at the bottom of the hole can be calculated using the closed in drill pipe pressure, known as Pg,. This must be the Pap recorded after the first build up and not a higher value that might result from a secondary build up; i.e. we assume that bottom hole pressure is equal to the formation pressure. Thus : P,=BHP=Pg)+Dp, (1) where P, = pore pressure in kPa (psi) Pap = final closed in DP pressure in kPa (psi) p: lensity of drilling fluid in use in kPa/m (psi/ ft) D = true vertical depth of hole in m (ft) Note that the right hand side of equation (1) will always give the bottom hole pressure but that this may not be equal to the formation pressure. A similar equation can be written for the annulus as well. This contains one unknown, the density of the influx, thus : BHP =Pagg +(D—bints)P1 + Bigg XPinci 2) (WCF Training Page 71 v1.00where Pann = final closed in annulus pressure in kPa (psi) hia = the true vertical height of the influx in m (ft) Ping = density of influx in kPa/m (psi /ft) Equations 1 and 2 can be combined and rearranged to give Pap + P1D = Pann +(D— Bing )P1 + Binet X Pint Ping = P1-— (3) Rise 1.5.3.3.3 Influx type Although the type of influx does not change the killing procedure, a gas influx normally increases the difficulty of killing a well because temperature and pressure changes have much greater effects on gasses than on liquids. In consequence, for the same initial closed in pressure, a gas kick will cause higher surface pressures during the killing process and will also cause higher pressures at the casing shoe until the top of the influx reaches it. For these reasons it is good to know what to expect. The influx type is determined by calculating the density of the fluid that has entered the well bore. Referring to Figure 1.15, the procedure is as follows: * Determine the volume of the influx Viag (see above) * Calculate the true vertical height the influx occupies in the annulus hing. (to calculate the height it is necessary to know the DC/OH and DP/OH capacities) Calculate the gradient of the influx pian from the equation (3) above. Compare the calculated ping with typical gradients for gas, oil and water. Typical ranges of gradients are: gas / condensate 3-0 kPa (0-075 - 0-15 psi/ft) mixture of salt water and gas or oil : 3-0- 9-7 kPa (0-15 - 0-47 psi/ft) salt water 19-7 - 10-7 kPa (0-47 - 0-52 psi/ft) 1.5.3.3.4 Selection of new drilling fluid gradient The eventual goal of all secondary well control operations is to regain primary control. In the case of a kick caused by drilling into an over pressured forma- tion, this will require that the well is circulated to a drilling fluid of a higher density, Again, the closed in drill pipe pressure is used to calculate the minimum density that will be required, as follows : . ie =p,+—2 P2=Pit—y where : Pp = density of new drilling fluid in kPa/m (psi /ft) Page 72 IWGF Training V 1.00This will be the fluid gradient first used to kill the well. Once the well has been killed, the density of the drilling fluid will be raised further to allow for a trip margin. 1.5.4 REMOVAL OF INFLUX 1.5.4.1 REVIEW OF METHODS 1.5.4.1.1 Circulation Provided there is a string of pipe in the hole and the influx is above the bit then the easiest and fastest way to remove the influx is to circulate it out via the choke or back pressure manifold (see Figure 1.16). A variable choke is used to exert back pressure on the drilling fluid as it leaves the annulus and in this way BHP is maintained at or slightly above P,. Figure 1.16 : Choke and kill manifold ‘There are two main methods for circulating out the influx. These are known as the “Wait & Weight” method and the Driller's method. In the “Wait & Weight” method, the well is killed in one circulation as drilling fluid of the new required density (see above) is mixed before starting and is used to displace the influx. In the Driller’s method, the influx is first circulated from the well using the original drilling fluid. A second circulation is then carried out to displace the well to fluid of sufficient density to balance the formation pressure. The advantages and disadvantages of the “Weight and Wait“ method compared to the driller's method are summarised in Table 1.5.1. Both methods will be discussed in detail later. IWGF Training __ Page 73 v1.00 @‘Table 1.5.1 : "Walt & Weight" vs. Driller's method ‘Advantages Disadvantages ‘* Surface pressures are lower during later stages of the kill due to the presence of heavy fluid in the annulus. ‘The maximum pressure exerted on the casing shoe is sometimes lower. This occurs if kill fluid enters the annulus before the top of the influx reaches the shoe. The well is killed more quickly since only one circulation via the BPM is required. This reduces the amount of time that the surface equipment is under pressure. 1.5.4.1.2 Migration ‘* The method is more complex to manage during execution. In the drillers method, circulation can be started without making calcu- lations (except those required for calculating the drill string capacity) Ensuring that the drilling fluid pumped into the well is weighted up correctly and that a constant density is maintained can be problematic. With the drillers method, pumping can begin as soon as stabilised drillpipe pressure is established. This could be important in case of a low MAASP in combination with a gas influx. Sometimes it is not possible to circulate the influx out of the well. An example of this is an influx which occurs when cementing a casing string in place. Another example is an influx which occurs when the drill string is not in the hole. When the influx is gas, one choice is to allow the influx to migrate up the annulus while controlling the surface pressure (and allowing the influx to expand) to prevent a further influx or formation breakdown from occurring. Once the gas reaches surface it is bled off and replaced with drilling fluid. This is generally known as the Volumetric method. If the influx is a liquid this method is not applicable as the influx will only migrate very slowly, if at all. A variant of the volumetric method is the Combined Volumetric and Stripping Method. This can be applied when the bit is above the influx (or even when the string is completely out of the hole) as is usually the case when a kick occurs while tripping. In this case, simultaneously with control- ling surface pressures according to the principles of the volumetric method, the string is stripped back into the well through the closed BOP. Once the bit is back on bottom, the influx can be circulated out. 1.5.4.1.3 Injection Sometimes it may not be possible to apply the other standard techniques or it may be preferable not to allow the influx to come to surface at all. Examples Page 74 IWGF Training V 1.00of such situations are when the influx contains a high concentration of hydrogen sulphide or when a gas kick is taken in a deep high pressure well which would result in excessively high surface pressures. In these cases it may be attempted to inject or bullhead the influx into the weakest exposed formation. Whether this is possible or not will depend on what additional pressure is required to inject the influx, especially if it is first necessary to squeeze drilling fluid into the recipient formation. Great attention must be paid to checking that the squeeze pressures are within the burst and collapse ratings of the well equipment and that formation fractures which cause addi- tional problems will not result. Bullheading is, for these reasons, usually only considered in special circumstances. 1.5.4.2 CIRCULATION The following discussion will be based on the “Wait & Weight” method. The differences for the Driller’s method will be discussed separately together with the circumstances when it may be preferable to apply it. Furthermore we will initially discuss vertical wells only. Deviated wells require the application of the same principles but the calculations are somewhat more complicated due to the necessity to differentiate between true vertical depths and along hole depths. Pressures are calculated mainly with the former while the latter are used to calculate volumes. 15.4.2.1 Phases There are several critical stages during a well killing process. Four of these are recognised in particular and are referred to as the four phases in well control. These phases (see Figure 1.17) are as follows for the “Wait & Weight” method: Phase = the time to displace the contents of the drill string to py drilling fluid. Phase II = the time from end of phase I until the top of the influx is at the choke. Phase III = the time from the end of phase II until all the influx is out of the hole. Phase IV= the time from end of phase III until all the p, drilling fluid, which was originally in the string and is behind the influx, has been removed from the well bore. The length of these phases is often measured in pump strokes rather than volume directly or time. This is convenient since when killing the well the pump stroke counter will be used to determine how far the kill has proceeded. The stage where the volume pumped places the influx inside the casing shoe (the shoe line), is important as well but has not been included in these basic phases in well control. The reason for this is that the occurrence of this stage could happen before or after the end of phase I depending on the distance the influx has to travel before it enters the casing shoe. IWGF Training Page 75 V 1.004. Phase | . Phase Il ©. Phase Ill 4. Phase IV Figure 1.17 : Four phases in well control (Wait & Weight method) 1.5.4.2.2. Circulating pressures While proceeding through the four phases just described, the well killing process is controlled by reference to the static drill pipe pressure (P,) and dynamic standpipe pressure (P,) at the top of the drill string. The definitions of these terms will become clearer in Topic 1.5.4.2.3. Again (see Topic 1.5.3.3.2) the drill pipe pressures are used because the contents of the drill string are accurately known and therefore these pressures provide an easily deduced relationship for bottom hole pressure. The slow circulating rates described in Topic 1.5.2.4 will give an indication of the circulating pressure drop to be expected at the anticipated killing pump speed. It is nevertheless important to be able to estimate the effects of a change of circulating rate or a change of drilling fluid density. It is explained in the WEDLP, Topic 2 of Section 2, Part 2 - Hydraulics that while pumping through a fixed conduit the pressure drop (P,) depends on the fluid rheology and the volume per unit time pumped (P, = CQ®). If all factors remain constant except the volume Q pumped then: Pe, = Pe, x (&) where: P, = the circulating pressure drop over the system Q =the pumping rate Page 76 IWCF Training v1.00In drilling operations it has been found that the power ‘n' varies. However, a value of 1.86 for an overall water based system is a generally accepted approximation. In addition it has been explained that if the density of the fluid changes but all other factors remain constant the pressure drop changes approximately as Pp oP, x a relationship : 1 Pumping rates during killing jobs are normally maintained at almost half the volume used during normal circulation. At such rates and in conventional wells i.e. not slim hole wells, the pressure drop over the annulus is so small (S 10% of total pressure loss) that it is negligible and therefore the pressure drop to move the fluid through the system is mainly used to overcome the internal string friction losses. The three pressures mentioned here are related thus: Pa =Pap + Pe In words this simply states that the overall circulating pressure or standpipe pressure equals the sum of the drill pipe pressure (due to the underbalance of the fluid column with respect to bottom hole/formation pressure) and the circulating pressure drop in the system at that circulating rate. 1.5.4.2.3 Static and dynamic pressure plots ‘The concepts of the four phases and the circulating pressures are combined to prepare a pressure v/s volume schedule to monitor the killing process. The simplest way of representing that schedule is on a graph. Phase I of the kill is the only phase where P,, will vary, due to the combined effects of the higher density kill fluid reducing the underbalance with respect to the formation pressure and simultaneously causing an increase in the circulating pressure drop. At the beginning of phase I just after starting the pump, and providing that the choke is properly handled (see later in Topic 1.5.4.2.5, where the start up of the pump is described), the initial standpipe pressure P,,, will be given by: Pp, =Pap+Pe, Pip is the initial closed in drill pipe pressure measured at the end of the first build up. P,, is the circulating pressure drop at the pump speed employed for the Kill with the original drilling fluid density. At the end of phase I, the new standpipe pressure Py, will be equal to the circulating pressure drop with the new drilling fluid density only, P., IWCF Training Page 77 v1.00Pgp no longer appears as it is now zero since the new drilling fluid density is sufficient to balance the formation pressure. The standpipe pressure should remain constant at P,., during the remainder of the well kill (phases I, III and IV) since the conditions in the drill string should not change and BHP should stay constant if the procedures are properly performed. Py, and Py, can now be plotted on a graph of pressure vs. strokes pumped. This is sometimes known as a kill graph and is shown in Figure 1.18. In a vertical or near vertical well, such as is considered here, the two points are usually joined with a straight line when the graph is drawn. Note, however, that this is a simplification as will be explained later. P,,, is held constant ~ from the end of Phase I to the end of Phase IV as explained. A similar graph of pressure vs. strokes can be plotted for the choke pressure. Pe, Pe, Pap ———> Pressure & Phase | Phase 1! Phase il Phase IV > _Volume/strokes pumped Figure 1.18 : Standpipe pressure graph Since the annular circulating pressure drop is generally negligible, the choke pressure will depend only on the difference between the combined hydrostatic heads of the fluids in the annulus and the required bottom hole pressure (i.e. BHP equal to formation pressure). For the same reason, Paa,, which is the (static) pressure measured on the annulus when no circulation is occurring, is usually virtually equal to P.,, which is the (dynamic) pressure measured just up stream of the choke during circulation. The choke pressure graph is more complicated to draw despite the fact that the circulating pressure drop is neglected because of two factors : Page 78 IWCEF Training v1.00The geometry of the annulus changes several times throughout its length. This means that the height which a given volume of fluid occupies changes as that volume moves up the annulus. An example of this is that the height occupied by a liquid influx initially is quite great since it occupies the space between the drill collars and the borehole. As soon as circulation commences, the influx will move into the space between the drill pipe and the borehole which has a greater capacity. The influence of the influx on the choke pressure is therefore reduced. In the case of a gas influx, the volume of the influx changes as it is circulated to surface. The influx must be allowed to expand, otherwise the gas pressure would remain equal to that at, which it entered the well bore. This follows from the laws of Boyle and Gay Lussac which can be combined to give PiV, _ P2Vo TY OT} where : P = the pressure of the gas V = the volume of the gas T = the temperature (absolute) of the gas and it is assumed that the compressibility is unchanged > Pressure Phase | Phase it Phase Ill Phase IV \Volumefstrokes pumped — Figure 1.19 : Choke pressure graph with gas influx An example of a choke pressure graph is given as Figure 1.19. In general such. a graph would not usually be manually constructed during a real well killing situation due to the time taken to do the calculations. Many well control software packages can however produce such a graph very rapidly in which case it forms a useful additional reference to check that the kill is proceeding according to plan. The general equation for calculating P,,, at any point is as follows : IWGF Training Page 79 v1.00 @P, ‘ann 7o—(D—hintr ~ hy, ).P1—Bp, -P2 — bine Pint where: hp, = the height of the column of ps fluid (will be zero until the end of Phase I) Pann = the annular surface pressure hige1 = the height of the influx the volume of influx divided by the annular area in which it resides and multiplied by the cosine of the hole angle The compressibility effects of gas mean that calculating the height of the influx once it has moved up the annulus is not straightforward. Formulae for approximating the pressures in gas influxes once they have moved up the annulus are covered in section 2.3.4.16 of EP 89-1500, the Pressure Control Manual for Drilling and Workover Operations. They will not be covered farther in this Part. 1.5.4.2.4 Selection of circulating rate Pump rates selected for well killing operations are normally half or less than the rate used for drilling. This is for a number of reasons but the most impor- tant is that the slower pump speed facilitates operation of the choke to follow the standpipe kill graph. This is especially the case when it is necessary to. cope with wildly fluctuating pressures when a gas influx reaches the choke manifold. A higher pump speed can mean that things happen too fast for the person operating the choke to react in time. Secondary limits on pump rate are ensuring that neither the surface gas handling capability of the mud gas separator nor the barytes adding capa- bility of the equipment on site are exceeded (see 5.2.5 above). The following example demonstrates how the amount of barytes to be added to the drilling fluid can be calculated and thence the maximum pump rate determined for a given rig mixing capability. Density of the drilling fluid in use: 11-00 kPa/m (0.486 psi/ft) Required fluid density to kill well: 13-00 kPa/m (0.575 psi/ft) Weighting material mixing capacity: 350 kg/min (772 Ibs/min) Double acting duplex pump capacity: 30 Vstroke (0.1887 bble/stroke) ‘The amount of weighting material per unit volume of p, drilling fluid is given by: ___4y x(p2-91) N= 5:815% dy x(P2 -P1) (9-81x10 xd,,)=p2 O%ag)~ Po Page 80. IWCF Training V 1.00where: N = the amount of weighting material kg/m _—_(Ibs/bbl) d= the density of the weighting material kg/m? —(Ibs/ft3) pi = the original fluid density kPa/m (psi/ft) 2 = the required density to kill the well kPa/m (psi/ft) dl For barytes with a density of 4,200 kg/m§ (262.3 lbs/ft ) this can be simplified to: yy = 3:815%262.3%(p»—~p) “CBE BL pa = 41200x(18-0- 11-0) yy - b472-8x(0-575- 0-486) ~~ (41-20- 13-0) a (1-822-0-575) N = 2978 kg/m? N= 105-1 Ibs/bbl ‘The maximum flow rate is given by: nn, = Weighting material mixing capacity Weighting material required ‘Thus _ 350 | ay: Erma : Qmax = 397g 71275 m/min Qanax = {pg 7 77845 bbls /min This would allow a pump speed of 1,175/30 = 39 spm (7-34/0-1887 = 39 spm) ‘The addition of large quantities of barytes to the drilling fluid will cause an increase in the total volume of the active system and sufficient room must be made available in the tanks to accommodate this. 1.5.4.2.5 Circulating out the influx Following the steps outlined above to select the well killing parameters, it is time to apply them. The procedure for the first circulation is as follows; * Before starting the killing operations, bleed off at the choke any increase in Pup pressure resulting from a second build up (see "monitoring surface pressure and volume” on page 70) to, or near, the initial stabilised closed in pressure. (Le. the closed in pressure after the first build up): * Open the choke and simultaneously start pumping the p, drilling fluid, bringing the pump rate to the selected killing pump speed. * While reaching and maintaining the pre-selected pump speed adjust the choke opening until the choke pressure P,, equals the closed-in annulus pressure P,,,, immediately before starting the pump. (Record the choke pressure throughout the first circulation.) IWCF Training Page 81 v1.00* Read the standpipe pressure Py, . It should agree with the calculated value, i.e. the normal pre-kick circulating pressure drop P,, at the selected pump speed plus the closed-in drill pipe pressure Pj, If the observed standpipe pressure does not agree with the calculated value, consider the observed pressure to be “correct” and calculate the actual Ps, . The revised P,, should then be used to calculate a corrected P,, and the standpipe kill graph then redrawn. * Continue pumping at the pre-selected killing speed and keep the standpipe pressure in line with the calculated pressure (the sloping line of phase I on the kill graph and thereafter the horizontal line of the py drilling fluid circulating pressure drop P., ) by opening and closing the choke as required. * Continue pumping with P,, equal to P., until the influx is circulated out. Be prepared to cope with a substantial increase in total surface volume of drilling fluid due to a gas influx expanding. * When all the influx and the original p, drilling fluid from the string has been circulated out ive. at the end of phase IV, stop the pump and close the choke. Check for pressures on the drill pipe and annulus gauges. They should both read zero. If not and if Py, = Pago, the choke may have been closed a little too quickly trapping pressure from the pump. Bleed off a little pressure from the annulus, checking that both pressures drop and do not rise again once the choke is closed. Open the BOP and make a flow check. Some imbalance between the annulus and drill string is likely, but a definitive flow can usually be confirmed. If a positive flow is still apparent, close the BOP and continue circulation under controlled conditions i.e. via the choke. + If there is no flow, start the pump and check whether the string is free. If free the string should be moved at regular intervals until ready to pull out. * Make a second circulation raising the drilling fluid gradient again to include a trip margin and conditioning the fluid to remove the effects of any contamination from the influx. While circulating during the well kill, the following actions should also be carried out : * Maintain and record the density of the drilling fluid pumped into the drill string. Ensure that it has the correct value * Measure and record the properties of the drilling fluid returns De-gas, treat, or separate for disposal, any contaminated drilling fluid returns . Page 82 IGF Training v1.001.5.4.2.6 Driller's method There will be situations in well control where the “Wait & Weight” method can not be used or where there may be a good reason to first circulate out the influx with the drilling fluid in use before then killing the well with a higher drilling fluid gradient. This method is called the driller’s method and, as with the “Wait & Weight” method, requires that circulation at or near the source of the influx, is possible. In the driller's method the well is killed in two circula- tions. In the first circulation the influx is circulated out using the original drilling fluid. In the second circulation, when weighted-up drilling fluid is circulated into the well, primary control is restored. The driller's method may be preferred if there are reasons for circulating out the influx as soon as possible or if it will take too long / be difficult to mix kill fluid (e.g. insufficient barytes on site). The advantages and disadvantages of the two circulation methods are contrasted in Table 1.5.1 on page74. As stated there, one factor in deciding whether to use the “Wait & Weight” method or the Driller’s method will be whether or not the top of the influx enters the casing shoe before the contents of the drill string has been displaced. If it happens before the end of phase I there is no difference between the two methods where the maximum pressure exerted on the casing shoe is concerned and this is normally the most critical point in the well kill. If py drilling fluid enters the annulus before the top of a gas influx has reached the shoe it is possible that the moment of highest pressure at the shoe during killing does not coincide with the top of the influx being at the casing shoe. This is because the denser fluid in the annulus below the influx will allow BHP to be maintained at Py with lower surface pressures. First circulation. ‘The need to reduce pump rate to accommodate the weighting material mixing capacity of the rig is no longer valid. However, to prevent unwanted pressure surges due to choke blockages and gas surges it is common practice to circu- late with approximately 50% of the usual circulation rate applied for drilling operations (just as in the “Wait & Weight” method). The standpipe pressure at the start of the first circulation is the same as with the “Wait & Weight” method. The standpipe pressure should then be approxi- mately equal to the normal pre-kick circulation pressure at the selected pump speed, plus the closed-in drill pipe pressure. Since there is no change in the gradient of the fluid being pumped in, however, the initial standpipe pressure must be held constant throughout the first circulation to ensure that the bottom hole pressure is also kept constant at, or slightly above Po. The procedure for the first circulation is as follows: * Open the choke and simultaneously start pumping the p drilling fluid at the selected pump rate. * Whilst reaching and maintaining the selected pump rate adjust the choke opening until the choke pressure equals the closed-in annulus pressure. IWCF Training Page 83 v1.00 @Record the choke pressure throughout the first circulation. * Read standpipe pressure. It should equal the normal pre-kick pump test circulation pressure at the selected pump speed plus the closed-in drill pipe pressure. If the observed standpipe pressure does not agree with the calculated value, consider the observed pressure to be correct. * Maintain constant standpipe pressure and pump rate whilst the influx is circulated out. * When all influx has been circulated out, stop the pump and close in the well to check the closed-in drill pipe and annulus pressures. At the end of the first circulation the closed-in pressures of the annulus and drill pipe should be the same and equal to the initial closed-in drill pipe pressure. The well is controlled but not killed. © Weight up spare drilling fluid to the required density Second circulation. Once the drilling fluid has been weighed up to the correct density the second circulation can begin. This is carried out in exactly the same way, including construction of the standpipe kill graph, as the first circulation of the “Wait & Weight” method. However, since the influx has been displaced with p, drilling fluid during the first circulation, large fluctuations in fluid returns and, there- fore, choke position are not expected. Thus, if possible, the density of the drilling fluid in the well is raised directly to that required to resume normal operations including a trip margin. During circulation the choke pressure Py, can be expected to remain constant until p, fluid reaches the bit, and will fall thereafter as the new fluid rises in the annulus. If all goes well the choke pressure will be zero at the time when the p> fluid reaches the choke. 1.5.4.2.7 Deliberate safety margins during circulation The principle of keeping bottom hole pressure equal to the formation pressure rather than higher is usually not adhered to in practice. Additional deliberate safety margins are often employed, such as including an overbalance when calculating the gradient of the kill fluid and / or applying extra back pressure with the choke. Overbalance Ideally, when killing a well, the new drilling fluid density should be such that the pressure it exerts at the bottom of the hole exactly equals the formation pressure. In practice, a drilling fluid density fractionally higher than the one calculated is often used. This provides an overbalance on the formation pressure. This overbalance can reduce the time needed to achieve an uncont- aminated drilling fluid column, but should only be included when the forma- tions exposed are known to be strong enough to support the additional pressure. The effect of the overbalance is shown graphically in Figure 1.20. To have the correct pressure in phases 2, 3 and 4, the stand pipe pressure P,t2, Which is Page 84 IWGF Training v1.00equal to Pz, would have to be reduced by the amount of this negative pressure. This is not usually done, thus pressures on stand pipe and annulus are too high by this amount. When a small amount of overbalance is used this will not greatly increase the risk of a formation breakdown. However, it should be realised that at a depth of 8,000 m (10,000 ft) even an overbalance as small as 0-35 kPa/m ( 0-015 psi/ft) in fluid density would make a difference of 1,000 kPa (150 psi) in bottom hole pressure. Phase | Phases lf Ill and IV Figure 1.21: Overbalance during well killing Extra back pressure In some situations an extra back pressure on the choke is applied as a further insurance against additional inflow into the well bore. This procedure is not generally recommended. Any additional choke pressure applied must take into account the total pressure applied to the open hole and the known strength of the exposed formations. In general, the pressure applied to the hole should be limited to the essential minimum. When extra back pressure is used the calculated value of the standpipe pressure is raised by the size of that back pressure. To achieve the back pressure desired, the standpipe pressure is often reduced in steps rather than as one continual gradual change. This is illustrated in Figure 1.21. Whether to step down or follow the line will be a matter for personal preference. Stepping down can be more difficult to accomplish requiring more manipula tion of the choke. Often, once the correct choke opening is established the standpipe pressure will decrease as per the kill graph with little further intervention. The method of applying extra back pressure to add a safety margin is preferred to that of using a higher fluid density. This is because extra back pressure can easily be reduced at any time during the well kill. IWCF Training Page 85 v1.00Phases tl, and IV 2 Pressure > Volumelstrokes pumped Figure 1.21: Stepping down of standpipe pressure with extra backpressure but without overbalance 1.5.4.2.8 Hidden safety margins In the circulation pressure In addition to the deliberate measures mentioned above, there are hidden safety margins that result from simplifications made when constructing the standpipe kill graph. Circulation pressure losses can be divided into three parts : * annular friction loss ‘© drill string friction loss © pressure drop over the bit For simplicity during construction of the kill graph, these losses are assumed to increase proportionally with increasing drilling fluid density. This estima- tion is conservative: the increase in circulation pressure loss proves to be less. than expected on the basis of direct proportionality with increasing drilling fluid density. This assumption thus leads to a hidden safety margin. Annular friction loss During well killing, circulation rates are decreased and therefore the flow in the annulus is often laminar. In the case of laminar flow, an increase in density will actually have no influence on the circulating pressure drop. A typical 3,000 m well with a surface wellhead might produce an annular friction loss of the order of a few hundred kPa. Thus if the drilling fluid density for killing the well is 10 or 20% higher than the original density, the inerease in annular circulating pressure drop will be over estimated by 50 to 100 kPa and this will produce a safety margin of that magnitude. This is illus- trated in Figure 1.22. Note : In offshore wells with sub sea wellheads, the line between the BOP stack and the choke manifold is sometimes very long. This can have a signifi- cant effect on annular pressure losses and must be considered carefully. The techniques involved for dealing with long choke lines are dealt with in Section 7 - the offshore section - of the distance learning package. Page 86 IWF Training v1.00Fig (star Fe | 4071521 2 Pg (ond of Phase |) 1 Seer ‘Volumeltrokes pumped Figure 1,22: Safety margin in the annulus under laminar flow conditions Drill string friction loss The flow regime inside the drill string is generally turbulent. In the case of ‘turbulent flow the pressure losses increase with increasing drilling fluid density but less than predicted by assuming the increase is proportional to the increase in fluid density. This leads to a second hidden safety margin of similar magnitude ie. 100 kPa in a 3000 m well with 5" drill pipe and a fluid density increase of 10%. This safety margin is illustrated in Figure 1.23 Faq (start) 875/126: FF(end of phase 1) 47/165, WH Pressure Volumatstrokes pumped Figure 1.23: Safety margin in the dril string Bit pressure drop The difference in bit pressure drop caused by the drilling fluid density increase, is taken into account when calculating P,, , prior to constructing the standpipe kill graph. On the graph the transition from P,, to P,, takes place at the same rate throughout phase I. In reality, this extra pressure drop only occurs at the end of Phase I when the denser fluid actually reaches the bit. The calculated extra pressure drop across the bit, caused by the heavy fluid, causes an extra safety margin which increases proportionately during Phase I is lost at the very moment the heavy fluid reaches the nozzles. This is illustrated in Figure 1.24. Again, its magnitude is usually low at a few hundred kPa. IWGF Training Page 87 v1.00Pag (start) 1012/131 Figure 1.24 : Safety margin at the bit 1.5.4.2.9 Pressure differential due to non-proportional change In hydro- static head The drill pipe pressure is assumed to decrease proportionally with the pumped volume until its value is zero at the end of Phase I. This is only true if the inside of the drill string has a constant diameter. Usually there is a decrease in diameter from the drill pipe to the heavy wall drill pipe to the drill collars. The practical result of the simplification is that the denser Grilling fluid initially progresses more slowly down the drill string, with a corresponding effect on bottom hole pressure, than is predicted by the kill graph. Thus, to follow the graph, the choke pressure is decreased more than it actually should be. This could result in an under balanced situation (as illus- trated in Figure 1.25) but fortunately the pressure differential is of a similar size as the safety margins already discussed and they tend to cancel each other out. Preseure Difference in P when DP and HWOP filled with heavy mud ‘Volumeistrokes pumped Figure 1.25 : Pressure differential due to non-proportional change in hydrostatic head In practice the hidden safety margins and the pressure differential due to the non-proportional change in hydrostatic head can usually be ignored due to their small size and self-compensating nature. Well engineers must be aware of these effects, however, and understand them in order to recognise when Page 88 IGF Training V 1.00they ought to be considered. A particularly important example is when a tapered drill string, which uses more than one size of drill pipe (a combination of 5" and 31/2" drill pipe for example) is in use. The non-uniform internal diameter may then be significant enough to cause a second influx if not accounted for when constructing the kill graph. Kill graph for tapered drill string To construct the static condition Pyp graph, Pgp, at any point of interest, ‘x’, is caloulated using the following equation: Pap, = Pap ~{Dx(P2-Pa)} To address the dynamic situation, a simple modification to the Phase I P,, kill graph has been devised. The only extra information required is the depth of the crossover from large to small drill pipe. The two calculations required are: + the drill string volume, and * the dynamic standpipe pressure at the x-over point. Derivation of the dynamic situation equation is as follows: P,., = dynamic pressure or P,, +static pressure or Pap, For well control purposes the dynamic pressures are assumed to be the cireu- lating pressure losses in the drill string only. In the Phase I P,, kill graph calculations we assume a linear relationship between P,, and P., : P,, - Pz, = the total dynamic circulating pressure increase due to the py drilling fluid. In the larger drill pipe at the top of the drill string, the friction losses per unit length are less than that in the smaller drill pipe at the bottom . It is there- fore assumed that a proportional amount of circulating pressure increase oceurs per volume of py drilling fluid pumped : A proportional string volume at depth 'x! can be written as VK Note: While the last two assumptions are not strictly true, the adjustment is simple and sufficiently aecurate for the situation, that is, appro- priate for field practice. ‘Therefore the dynamic pressure at depth ‘x’ can be written as: y. =Py, “{(P. Pega} Substituting the above in the equation for P,,, for depth ‘x’ equal to the crossover depth gives: IWGF Training Page 89 v1.00 @To Pressure Pu, = Pa, f(r. Pe gc} fDeto0-o0} Py,,,= standpipe pressure observed at start of kill (kPa or psi) a circulating pressure at reduced pumping rate (kPa or psi) circulating pressure at end of Phase I, (kPa or psi). V, = string volume up to crossover (m® or bbls) Vestal = total string volume (m? or bbls) Dg= depth of crossover (m or ft) P1 = original drilling fluid density (kPa/m or psi/ft) P2 = kill drilling fluid density (kPa/m or psi/ft) construct the kill graph four points of interest are calculated: © Py at start of kill * Volumes V,,,.) and V,, (the latter for each section of drill string or BHA with a different inside diameter) oD won * Py, at the various crossover points using the above equation * P,, at the end of Phase I Now mud enters smatier bore dill pipe. Figure 1.26 : Kill graph for a string with two ‘Volume’strokes pumped sizes of drill pipe and BHA Page 90 IGF Training v1.001.5.4.2.10 Deviated wells ‘Until this point we have considered vertical wells only. However, many wells are now drilled with considerable deviation. When well control situations are encountered in these wells it is important to differentiate between along hole and true vertical depths to produce an accurate standpipe kill graph. The equation used for a tapered string is suitable and can be very easily adapted to be used to calculate Py, at any point of interest, x, in a deviated well. The equation in its general form is: P, _ “(le Pa) Fe} Poe, x(p2—pa)} If only one size of drill pipe is used, thon V,/Viotai becomes Dap, /Dan,., Since drill string volume is the product of length times capacity and if the capacity is constant for the whole string (neglecting the small error made by omitting the drill collar capacity difference and the pressure drop over the bit), the equation then becomes: P, Patan +(e. Paget} fea x(e2-01)} Several Pj, values at various points of interest can be calculated for a deviated well. These points would most logically be: © the kick-off point, ‘+ end of the build-up section, * the start of drop-off section, and ‘© the end of drop-off section. The kill graph is constructed in a similar fashion as for the tapered drill string graph. The corresponding volume of pz drilling fluid, the statie Pay, and Px, at the various points of interest are calculated. These points are plotted and the appropriate straight linos connecting them are drawn. The error in Py, incurred by neglecting the true directional nature of the well results (depending on the well design) in an overbalance or an underbalance to the formation pressure until the kill density drilling fluid has reached the bit at the end of Phase I (see Figure 1.27). For highly deviated wells, this additional overbalance can be large enough to cause lost circulation or stuck pipe. The underbalance could lead to a second influx. In the graph for a vertical well, a straight line is drawn from Py. to P.,. ‘This is however not a correct representation for a deviated well. Consider the situation for a simple single build and tangent section well. If the vertical well straight Py, line was followed instead of the actual deviated well Py, line, an overbalance to formation pressure would occur in the vertical, the build, IWCF Training Page 97 v1.00