Vacuum Hold Time
Table 1 Vacuum Hold Time
Suggest retaining the existing vacuum hold time guideline in C57.93 but with
Additional periodic heating cycle to address the heat of vaporization issue.
• C57.93 is more stringent Reason: transformer insulation is greatly affected by moisture
• No further benefit if > 24 hours content and therefore we need to ensure a good dry transformer.
- “free water @ room temperature freezes @ 4 torr”
- “ heat of vaporization would lower the temperature to a point where remaining water
will freeze making extensive vacuuming pointless”
• Difficult to define “open time”, therefore hard to quantify T and/or t.
• Suggestion, use dew point to judge vacuum hold time.
Vacuum Hold Time
Table 1 Vacuum Hold Time
Suggest retaining the existing vacuum hold time guideline in C57.93 but with
Additional periodic heating cycle to address the heat of vaporization issue.
Reason: transformer insulation is greatly affected by moisture
• Roger Hayes:
content and therefore we need to ensure a good dry transformer.
Table # 1
• I can agree with leaving this table as is even though I believe 12 hours for 69 kV is more than adequate, however,
we need to somehow address voltages < 69 kV were 12 hours is too long and introduces costly processing at the
site. I would suggest 8 hours for < 69 kV.
• As vacuum time should be continuous, then any “periodic heating cycle” may have to be by short circuit current
application, which may not be practical in the field.
• Mike Lau – Heating cycle could be achieved by filling with hot oil to heat up the winding and draining it. This could
be repeated several times.
Phase Diagram of Water
Hold time and Set time
For New units For old and reprocessed units
Table 2 Table 4
• No reason for Old Units to take longer time for impregnation – reverse could be true
as it depends how long the core and coil is left without oil
• Suggest Table 2 and Table 4 should be combined into one table
Agreed, These two tables should be resolved into one.
Hold time and Set time
For New units For old and reprocessed units
Table 2 Table 4
• Roger Hayes:
The tables should be combined into one, however, if reprocessing is required for reasons as sited in clause 4.10,
(Note: 4.10 is on Reprocessing) then the higher times should be followed.
• For 69 kV. 48 hours is excessive. I suggest 24, which is still longer than I have used over the years. For < 69 kV,
these times could be reduced.
Agreed, These two tables should be resolved into one.
Hold time and Set time
For New units
Table 2
• Doble has different criteria for units with and without pump.
• Doble specifies time for running pumps
• Suggestion to run half of the pumps for 2 hr and other half for another 2 hrs.
Why not run all the pumps at once?
Yes, we should include the recommendation of running the pumps. ( why not running both pumps
at the same time??) Also, change time for 69KV to 24 hours?
Hold time and Set time – Running Pumps
• Kipp Yule: “A cautionary statement regarding static-electrification should be included regarding running all or many
oil pumps with cold oil, dry oil, and high velocity oil velocity. This is also called Streaming Charging, which is well
explained in the NYNAS OIL Handbook chapter 3-18. If Oil Pumps are used then the recommend the
electrostatic charging tendency (ECT) be measured & monitored. Also, known as:
•
Mike Lau:
I read up Streaming Electrification in the Nynas Transformer Oil Handbook. It is on Section 5.17 (P.109 - P110 )
on my copy.
• It describes the problem is more likely to occur when the oil has a flow speed exceeding 0.5 m/sec. (I don't think
this number means much to most users. Not sure if this could be roughly converted to the capacity of the oil
pumps ).
• It also indicates that well refined and inhibited oil would have lesser streaming charging.
Paul Mushuill:
I am also concerned about running pumps for two hours. I do not think that we can blindly state to run pumps
without some precautions. If you are in Phoenix or Florida, it is not a concern but most of us are going to have
much cooler oil.
Yes, we should include the recommendation of running the pumps. ( why not running both pumps
at the same time??) Also, change time for 69KV to 24 hours?
Vacuum Level
Table 3
• Suggest 2mm Hg for 69 KV; 1 mm Hg for 230KV; and 0.5 mm Hg for 345KV and above.
• Suggest Table 3 to be tied in with Table 1 ( Vacuum hold time )
• Suggest to use Total Dissolved Gas : <5000 ppm; or 2500ppm for conservator type
• Suggest to use Piper Chart to determine the vacuum level in order to achieve 0.5% moisture
content.
• Roger Hayes:
I can agree with your suggestion, despite believing that 1000 microns at 69 kV is more than adequate, however for
< 69 kV it should be 2000 microns
Suggest No Change.
Soak Time
• Should not provide two sets of criteria –”everybody will go for the minimum !”.
• One suggestion : all 24 hours
• One suggestion : > 325kV 12 hrs; < 325KV 4 hrs
• One suggestion : >230KV, 72 hrs; <230KV 24 hrs
Suggest 69 KV and below – 8 hrs, > 69KV – 12 hours
Soak Time
• Roger Hayes: I agree with your suggestion if it applies to “minimum energizing period”, however, although
somewhat subjective, I have always recommended 24 hours if there is no time penalty.
• Paul Mushill: My biggest concern is the energized at no load soak time. When a transformer is ready to go and
energized, our normal commissioning usually takes 8 to 16 hours. After that, operations is pushing to apply load
and put in service. If we have to wait a day or two, our testers have to come back out to do load checks after
applying load. I do not see any benefit to this soak time over the normal 8 hours commissioning time. If more soak
time is required, I would like for some manufactures’ representatives explain the benefits and purpose and
convince us that this is really necessary.
Suggest 69 KV and below – 8 hrs, > 69KV – 12 hours
• Should this table be included? (Looks like Piper Chart)
Is this the Piper Chart? Table does have some good information.
Should we also include 0.3% moisture?
• Is this necessary?
• “Concern potential for creating suspended bubbles in the oil resulted from residual gas”
Suggest don’t include
Other Comments
• C57.93 should provide a definite dryness acceptance criteria
e.g. Doble PF ? Moisture Content ? Dew Point?
• Confusing units: mm Hg, torr, inches of Hg, Pascal,
• Include conversion table
• Use limits of Dissolved Gas – to minimize amount of oxygen
• Update Maintenance sections
1. There will be sensor available to measure moisture in paper directly.
2. Yes, will include a Conversion Table.
3. Yes, use C57.106 guideline for total dissolved gases
4. Yes.
Other comments: Cold Trap
• Kipp Yule:
For field dry out of large units, I find it is useful to use a cold trap to measure the rate of moisture removal rather
than just a time based process. If the operator knows the weight of the insulation and the moisture content at
start, then the quality of moisture to be removed and be determined and measure. Usually needs some OEM
details to have accurate insulation weight.
• While the above is not new, it is only offered from the point of view if it should be included in C57.93.
Mike Lau:
As for the cold trap. My understanding that they offer very efficient and fast removal of moisture vapour:
• Here is the quote from Baron USA:
• One pound of water becomes 100 cubic feet of water vapor at 100 Torr.
• • One pound of water becomes 10,000 cubic feet of water vapor at 1 Torr.
• • A cold trap condenses that vapor into a solid ahead of the vacuum pumps.
• • That vapor then does not have to be removed by the vacuum system.
• • That water can then be measured.
Alan Peterson:
We have had good results using cold traps. They are easily installed in the vacuum line. We fabricated ours in our
shop from aluminum with a double wall container. We use liquid nitrogen for the coolant. The frozen moisture is
scraped out of the inner container and allowed to melt and be measured. On the subject of oil electrification, we
have found that different designs of oil pumps can contribute to this problem. Pumps designed with the output
directly entering the tank or tank gussets have a greater propensity for this phenomenon than pumps that have the
output connected to the tank or tank gussets through a 90 degree elbow.
• I can only surmise that this is caused by allowing the oil to remain in greater contact with grounded metal parts for
a longer period of time before entering the main tank.
