‘The Applicability of the Lugeon Method of Packer Test Analysis to Hydrogeologic Investigations ‘Timothy R. Roeper, William G. Soukup, and Robert L. O'Neill ECKENFELDER INC. Rochester, New York Mahwah, New Jersey ABSTRACT ‘The Lugeon method of packer testing and analysis was originally designed to assess the need for foundation grouting at dam sites and has been modified over the years to its present form as described by Houlsby (1976). The method consists of calculating Lugeon values for each of five tests conducted at increasing (P1, P2, and P3) and then decreasing (P4 and PS) pressures, followed by the interpretation of the pattern of results, and subsequently the selection of a representative hydraulic conductivity. This approach is intended to provide information about the formation itself and the potential for test pressures to alter the natural hydraulic conductivity of the rock. ‘The Lugeon method was required to be used at a site in western New York at which 1 total of 742 packer tests were conducted in fine grained sandstones and shales. ‘This paper presents a post-investigation evaluation of the cost effectiveness of thie method over the more conventional one-pressure approach. The results indicate that the Lugeon derived hydraulic conductivity was, on average, very similar to the value derived from the P2 pressure. These results, coupled with the associated cost of the additional tests, suggest that the single pressure test approach would be more cost effective. Furthermore, the data demonstrates that the higher pressures have the ability to locally alter the natural permeability of the rock. HISTORY ‘The Lugeon method of packer testing and analysis was originally designed to assess the need for foundation grouting at dam sites and has been modified over the years to its present form as described by Houlsby (1976). The "modified" test differs from the standard test (Lugeon, 1933) by the use of a range of lower pressures. Lugeon's standard test specified a consistent pressure of 10 bars. The "modified" method 661consists of five consecutive tests, each of ten minutes duration, completed at first increasing pressures and then at decreasing pressures. ‘The pressures used in the tests are related to depth by the following relationship: Low pressure (P1 and PS) in psi = 0.4 x depth of the test interval in feet; ‘Medium pressure (P2 and P4) in psi = 0.7 x depth of the test interval in feet; Peak pressure (P3) in psi = 1.0 x depth of the test interval in feet; ‘The five consecutive tests are completed in the following sequence: First ten minutes at a low pressure (P1) Second ten minutes at a medium pressure (P2) ‘Third ten minutes at a peak pressure (P3) Fourth ten minutes at a medium pressure (P4) Fifth ten minutes at a low pressure (P5) ‘A single Lugeon value is than calculated for each one of the five tests using the following formula: Lugeon value = water taken in test (i in) x 10 bars ___ augeon vE test (litres/meter/min) x Test pressure (bare) ‘The relative magnitude of the five lugeon values is then compared to the patterns shown on Figure 1 and a decision is made as to which of the five hydraulic conductivity values (or combination thereof) is most representative of the test interval. A brief discussion of the original interpretations of these patterns Houlsby, 1976) is presented below. Group 1 - Laminar Flow Laminar flow is assumed to occur when the five calculated lugeon values are all about the same. The reported representative permeability is taken as the average of the five values. Group 2 - Turbulent Flow Turbulent flow is interpreted when the lugeon value calculated for the peak pressure (P3) is less than that calculated from the two medium pressure tosts and when values for the low pressure tests are approximately equal. The representative hydraulic conductivity is taken as that calculated for the peak pressure. Group 3 - Dilation ‘When the lugeon value for the peak pressure (P8) is greater than those for the two low pressure tests, and when the two low pressure values are approximately equal, temporary dilation of the rock mass is inferred. The interpretation that the dilation is temporary is based upon the return of the lugeon values at the end of the test (P4 and PS) to similar values obtained at the beginning of the test (PI and P2). The representative hydraulic conductivity is taken as the average of pressures Pl and P5 662(6161 “heroH 0H) SISATVNY NOdON1 GaIsIGOW Tawnow (erwa nvm wo isan some anoarmaso S “ (Wash SVR KUATEDNANOST wave SRS (Gada SiPIVA AUALLDRGNOD Waive SAH ‘NOT do NOUWATUNLNI oso ‘MOT 4O NOLLVAURLNT ogo (©NLLNOYD) VIESLIHD SISATYNY NO39N1 663or alternatively, the average of P2 and P4 if the lugeon values for these pressures are less than those calculated for Pl and P6. Group 4 - Washout of Joint Filling Materials A progressive increase in the five lugeon values without a return to values recorded prior to introduction of the peak pressure (P3), is indicative of the permanent washing out of joint filling material or the permanent movement of the rock caused by the testing. This is indicative of the test pressures being too high. The representative value in this scenario is taken as that calculated from P5. Group 5 - Void Filling ‘A progressive decrease in the five lugeon values is regarded as an indication that the test is gradually filling empty voids, joints etc. or that the tested fracture is of limited extent. The recommended representative hydraulic conductivity for this scenario is calculated from P5. ‘A more detailed discussion of the interpretations, rationale, and applicability to grouting projects is presented by Houlsby (1976). STUDY MODIFICATIONS The lugeon method of packer testing was required to be used as a part of a hydrogeclogic investigation at an industrial site in western Now York. It was recognized prior to the start of the investigation that modifications to the interpretations made by Houlsby and other previous investigators would be necessary to reflect the difference in test objectives. As previously mentioned, the objective of these earlier analyses was to determine if grouting at a dam site is warranted. The magnitude of the Iugeon values would be used to assess the need for grouting whereas the pattern of relative values would be used to evaluate the type of grout best suited for the anticipated fracture type. If, for example, flow pattern was observed, it was interpreted that granular material was in the rock joints and thus the use of a chemical grout would be indicated. Turbulent flow would indicate large fractures and the need for a coarse grout, ete. By contrast, the objective of this investigation was to determine the natural hydraulic conductivity of the bedrock to groundwater flow under typical gradients. ‘This objective is not served by the selection of the P3 value in the turbulent flow pattern since natural gradients are seldom sufficiently high to facilitate turbulent flow. For this pattern, the Houlsby method was modified such that the geometric mean hydraulic conductivity of the two lowest pressures (P1 and P5) were calculated. The hydraulic conductivity calculated at these pressures is likely to be more reflective of the natural groundwater flow conditions. For the void filling sconario, the Houlaby method calla for the use of the final (P5) value since this will likely be the representative condition of the rock during the pressure grouting process. Under natural flow conditions however, these voids which are presumed to be above the level of saturation, will not be pressurized and 664thus the initial (P1) value would be most representative. These two modifications were incorporated to the interpretation of the lugeon patterns during this investigation and are illustrated on Figure 2. DATA ANALYSIS A total of 742 packer tests were conducted in fine grained sandstones and shales during the coarse of the investigation. These data were evaluated to determine the cost effectiveness and applicability of the lugeon method of packer test analysis as compared to the more conventional one-pressure approach. The data from each packer test were entered into a computer database which included the test location, depth, calculated hydraulic conductivity at each of the respective pressures (Pl through PS), the representative hydraulic conductivity value, and a criteria number which identified the lugeon group into which the test was categorized. Since there are five pressures, there are a total of 25 different lugeon patterns which are possible. A 26th category was established in which none of the pressures resulted in a measurable "take". The representative hydraulic conductivity value was than divided by the value calculated for each individual pressure. In those cases in which the individual value was greater than the representative value, the inverse of the calculated value was recorded. In this manner, the absolute value of the factor by which the selected representative hydraulic conductivity differed from the value calculated for each individual pressure was obtained. Thus, a perfoct match was represented by a factor of 1.0, and a difference of one order of magnitude was represented by ten, ete. Note that this relationship is not linear and these values cannot be used as conversion factors. They are simply used to provide a rough, order of magnitude, estimation of the difference between the calculated values. Once the above calculations were complete, the accumulated data were evaluated to determine the frequency at which each of the 26 categories were encountered, and the minimum, maximum, and average factors by which the representative hydraulic conductivity value differed from each of the respective pressures, This analysis was conducted on both the whole population, excluding those tests which had "no take" at the representative pressures defined by the lugeon patterns, and on subsets of the population representing the individual categories. Comparisons of the calculated hydraulic conductivities for Pl versus P5 were also made to evaluate the potential effect of the high pressure (P3) on the hydraulic conductivity of the surrounding rock mass. RESULTS A review of the data indicates that each of the 26 categories was represented by at least one of the 792 tests. However, over 86 percent of the tests were represented by one of only eight Iugeon patterns as listed on Table 1. The pressures used in calculating the representative hydraulic conductivity are parenthetically shown in the Table. The eight groups consist of the five modified lugeon patterns, a modified version of the dilation pattern, a ‘no flow" group, and a group in which there we "take" at P3 only. The "scattering" of the lugeon patterns is attributed to a variety 665Tana Ea eoruuonmameme EEG] sara pean amar aware | Es eave = nom sess | aa EEE] « Ea « oer 1201 — Gawoarvasa “ “ temas so wemasmeme a TA AAT aaa eae aa RIVA KEATS GROT ava — ea "MOT a0 NOMViMAcNaING wosoert ‘Ao do NoUvuaaealsa nosont (SNOLLVOLLSSANI DID010SDOUGAH) VINALIED SISATVNY NO39N1TABLE 1 ‘SUMMARY OF PACKER TEST DATA ‘AS DEFINED BY LUGEON CRITERIA LUGEON NUMBER PERCENT CRITERIA OF RECORDS OF TOTAL Laminar Flow (P1,P2,P3,P4,P5) Dilation (P1,P2,P4,P6) Turbulent Flow/Dilation (P1,P5) Washout (PS) Void Filing (P1) Dilation (P2,P4) Take at P3 only No Flow Other gkesssaas TOTAL & 667of causes including geologic factors such as fracture roughness, frequency, orientation, straightness, etc. Scattering of the lugeon values is reportedly common and can be seen in neighboring boreholes in fairly uniform areas (Houlsby). These observations were found to be consistent with the data collected during this investigation. ‘As discussed previously, the selected representative hydraulic conductivity was compared to each of the calculated values from the individual pressures for all tests except the ones at which "no flow " was recorded for the selected pressures based upon the lugeon patterns. ‘The total number of tests within this population was 415 and the resulting analysis is presented in Table 2. This population represented approximately 56% of the total population and indicates that the average difference between the representative hydraulic conductivity value as determined by the ‘Lugeon method and that determined by the individual pressures, ranged from a factor of 1.97 for the P2 pressure to 5.68 for the Pl pressure. Also noteworthy is that although the average factors for the P2 and P4 tests are very simil range, standard deviation and variance are lower for the P4 comparison, T' result of the "washout" group which indicates that the test pressures are too high. ‘The individual values obtained for these comparisons are illustrated in Table 3. The differences between the P2 and P4 comparisons for the "washout" group are attributed to the effect of the peak pressure (P3) upon the natural hydraulic conductivity of the rock. Considering the fact that whenever possible the most conservative (i.e. highest) representative value was chosen, the lower statistical values in the P4 comparison indicate an effective increase in the measured hydraulic conductivity for the P4 tests versus the P2 tests, Removing this conservatively biased group from the overall statistical comparisons results in an average, standard deviation, and variance for the P2 factors of 1.72, 2.05 and 4.25 respectively. The same statistics for the P4 comparison were calculated at 2.00, 3.16 and 9.96 respectively. ‘An additional indication of the effects of the peak (P3) pressure on the natural hydraulic conductivity of the rock was obtained by comparing the individual values obtained using pressure Pl and P2 versus P5 and P4. Additional comparisons were made as to the number of times at which the same tested interval went from a "no take" value at the initial low pressures to a calculated hydraulic conductivity at P4 and P5. ‘This analysis revealed that the hydraulic conductivity calculated at pressure P5 was greater than that calculated at pressure Pl on 178 occasions, or approximately 43% of the tests which took water at both Pl and P5. There were 29 instances where there was no take at P1 but take at P5. There were 180 instances in which the calculated hydraulic conductivity at Pa was greater than that at P2 and 5 cases where there was no take at P2 but take at P4. CONCLUSIONS AND RECOMMENDATIONS ‘The accumulated data indicates that, on average, the hydraulic conductivity calculated from the P2 pressure is within a factor of two times of that calculated by the Lugeon method. It is further apparent that the peak pressure (P3) has the ability to increase the natural hydraulic conductivity of the tested rock mass by 668TABLE 2 STATISTICAL SUMMARY OF DIFFERENCES BETWEEN THE REPRESENTATIVE HYDRAULIC CONDUCTIVITY VALUE AND THE VALUE CALCULATED FOR EACH PRESSURE (includes only tests in which there was # measurable take) ‘STATISTIC, REP, VALUE AEP. VALUE REP. VALUE AEP. VALUE REP. VALUE V.S.P3. V.S.P4 V.S. PS. MINIMUM 1.00 MAXIMUM 1126.19 AVERAGE 5.68 STANDARD DEVIATION 60.86 VARIANCE 3703.59 TOTAL NUMBER OF TESTS = 415 669wo ee wl ieoh ee. meee Oz ‘Ayuo eet 12 EL oer zz 00} 80% wh ISL we Ob (yei'Zed) von 69zb we 9h LL Bs LS weet Ly (ied) Bum POA, wee oe oot «peo OLE KELL 9p (Gd) snoysena ve6 90 cob 69h ye %¥9'6 =O (Gd'td) vonena/mol wenn, 190 0 00 6e0. zo | ort wero Lo (Sd'pd'd'bd) vont 80 VOL oo 290 wo | Set 0002 = £8 (Sd'pd'Ed'Zd" Led) MOL see] NOILVIASO NOILVIASO WIOL FBONVIUVA “ONVLS BONVIWA “ONVLS 40% |.1NNOO| ‘VIN3LIHO NOIOM vida Zales (oye oiqesnseow sem a1oyy Yorym Ut 5380) AjuO sopnjour) ‘vd GNY Zd SNSH3A ALIALONGNOS OFINVGAH AALLVINAS3Ud3Y HO VIHALIHD NOJDM1 4O AUVAWINS TWOLLSLLVLS es78vL 670either washing out any material which may be filling the fractures or voids and/or permanently lifting the overlying rock. Given the amount of time required to complete the Iugeon method, the subjectivity of assigning the lugeon pattern to a given category, and the above conclusions obtained from the evaluated data, a more cost-effective approach of packer testing for hydrogeologic investigations would be to use a single pressure, equivalent in psi to 0.7 times the depth, in feet, of the overlying geologic materials. In order to increase the sensitivity of the test in low permeability intervals, it is also suggested that the test be conducted for a duration of five minutes for those intervals at which a measurable amount of flow is recorded and ten minutes in very tight formations. It is estimated that the use of a single pressure test. as opposed to the lugeon method would have saved on the order of 250 hours of field observation time and 25 rig days for the drilling subcontractor on this Project, ‘The conclusions made regarding the use of the lugeon method for hydrogeologic investigations is further supported by the objectives of these investigations opposed to a grouting project. In most cases, one is looking for a conservative measure of the in-situ hydraulic conductivity. The data evaluated in this study indicates that the use of high pressures can increase the natural hydraulic conductivity of the rock mass and potentially provide erroneously high values. Although the method may indicate that a given fracture is limited in extent, this information does not justify the potential for artificially increasing the fracture permeability of the rock nor the amount of additional effort and resources necessary to conduct the additional testing. REFERENCES Houlsby, A.C., 1976. "Routine Interpretation of the Lugeon Water-Test," Quarterly Joumal of Engineering Geology, Vol. 9, pp. 308-313, 1976. 671