A88- How to keep the transformer dry during transport and installation- Part 2

In part 1 of this article, we have seen how to keep the transformer dry during transport preparation at the factory, transport to the site and storage in the field. A87- How to keep the transformer dry during transport and installation- Part 1 | LinkedIn

In this part, we will see how to prevent excess water ingress into the paper insulation of power transformers during installation work. 

7.   Installation

If sufficient care is not taken, an appreciable quantity of water can get inside the transformer from the outside atmosphere during the installation work. The following steps will reduce the water ingress into the insulation.

(1)   It is a good practice “to get the transformer active part under oil as soon as possible”. Plan installation work procedures such that exposure of active parts to outside humid air is minimized.

(2)   Transformer buyers may think that transport, storage and installation of transformers are in the scope of supply of OEM and they need not worry about these stages as the manufacturer has executed a two-year performance warranty for transformers. Some of them may also think that wetting the transformer from exposure is not a serious issue as hot oil circulation for a couple of days will bring back the insulation to the original dry condition.

Remember this high-value asset will give expected life only if you keep the factory dryness during installation work. It is almost impossible to remove moisture in paper by hot oil circulating for a couple of days. Transformers may work the guarantee period even with high moisture content, but faster paper ageing may lead to premature failure. So, users must appreciate that as soon as FAT (Factory Acceptance Test) is over or the trailer with the transformer enters the substation gate, it is their property, and they must have a close watch over the procedures to ensure the dryness of the transformer.

Nearly a quarter century back, we exported a 56 MVA 33/11 kV transformer from India to a US customer on the West Coast, that was for feeding power to Hollywood Studios. One day our service engineer, deputed for installation at the site, approached the station head announcing that he was going to open the transformer to fit the bushings. The station engineer asked him to wait. He checked the humidity level by taking out an instrument from his shelf. The station engineer refused permission to open the tank as the air humidity level was high. Our engineer was taken aback by the user's care for even a 33 kV class transformer.

(3)   When there is rain or the dew point of the atmosphere is high (RH> 70%), avoid opening the transformer tank. If the transformer is brought to a location warmer than the transformer itself, the transformer should be allowed to stand until all signs of external condensation have disappeared. When hand holes, manholes or turrets are opened out for installing bushings and other accessories, it is a good practice to feed dry air from bottles or a dry air-generating plant into the tank.

As far as possible, avoid persons going inside the transformer for active part inspection. Breathing and sweating from the skin can add to the moisture level in the paper insulation. When personnel are inside, feed breathable dry air (maximum dew point -45ºC) at a rate of 564 lpm (34 m3/hr) with an additional 144 lpm (9 m3 /hr) for each additional person for purging the tank.

Avoid opening more than one vent at a time in the tank to avoid cross circulation of outside humid air into the transformer. When any cover is opened, put back the blanking plate or cover the opening with a plastic sheet.

(4) When internal inspections of a transformer filled with natural ester insulating liquid are done, it is important to minimize contact with ambient air to prevent the oxidation of the impregnated natural ester liquid. The total exposure time to dry air shall not exceed seven (7) days if no alternative precautions are taken. Any process involving hot air circulation shall be avoided for these transformers.

(5) Wipe clean and remove surface moisture from all accessories going inside the tank e.g., bottom tail of bushings, inside of porcelain bushings, flow indicators and oil pumps.

(6) Moisture can get into the transformer through poor oil seals. During installation work, fresh gaskets shall be used to replace old gaskets wherever openings are made. e.g. manholes, hand holes, bushing turrets, and cooler pipe joints. Retighten all newly replaced gaskets after 24 hours. Take special care at the following critical locations. Many service failures were reported due to water entry through these joints with rubber oil seals.

• Top terminal seal in draw lead condenser bushings (improper placing of oil seal; forgetting to put back the seal after erection work)
• Oil seal of explosion vent and pressure relief devices
• Oil seals on the suction side of oil pumps: Wet air can be sucked through oil seals due to negative pressure created behind the pump.
• Breather pipe joints. With missed oil seals or defective gaskets, humid air will be breathed in through these joints, bypassing the silica gel breather and its bottom oil seal.
• The Connection between the bushing terminal pad and connector to the jumper wire shall be tight. With a loose connection, there are chances of overheating during transformer loading that may damage the top terminal oil seal, leading to water entry to the active part through the bushing tube.

(7) During assembly work, the transformer shall be sealed off after working hours. Apply slight dry air overpressure, less than 0.2 atm (20 kPa) for overnight keeping. The next morning, if there is a pressure drop, leaks are to be suspected. Leaks may be detected by leak detection instruments, with soap water or with a plastic bag tightened around valves  (will be inflated by leaking air) etc. In this way, the assembled seals may be checked successively as the work goes on.

8.  Leak Test

       Before the final oil filling, the tank is kept under vacuum for several hours to remove the surface moisture entered during the installation process and to eliminate air from the insulation. Leak tightness shall be checked at the start of the vacuum and again during vacuuming. If leaks are there, humid air can get in during the vacuuming process, wetting the insulation. Any air leakage into the transformer tank, while a vacuum is being drawn on the transformer, may wet the transformer insulation. When drawn into a vacuum, air expands and temperature drops, releasing moisture. If the core and coils are cold, the moisture released from the air will condense on these parts and will be absorbed into the paper insulation. To avoid this hazard, conduct a leak rate test before starting the vacuum processing. Normally this is checked as below:

• When vacuum levels reach 200 millibars (torr), switch off the vacuum pump. Close the valve and observe pressure rise inside the tank from leaking air. It shall be less than 0.5mm within half an hour.
• At 0.3kPa (2.3 millibars) vacuum, switch off the pump for 1 hour and then find pressure rise after 30 minutes (p millibars). Then the leak rate for the tank with an oil volume of V litres shall be less than 20 millibar litres/sec

            Leak rate (milli bar litres per sec) = p (millibar) x V (litres) / T (seconds)

• IEEE C57.93 standard for Transformer Installation, recommends a leak test as below: (Method B)

Take two readings of the pressure; 60 minutes and 90 minutes after the vacuum valve to the pump is closed, the first 60 min is allowed for the de-absorption of gasses from the insulation. A leakage rate may be calculated using the Equation,

Leakage rate = V (P90 – P60) divided by 30 minutes, where V is the oil volume of the tank in litres. It shall be less than 150 mm Hg L/ min (20 m3 Pa/min)

 9. Case Histories

Case histories of transformer failures due to water ingress during installation:

(1) Case history: 1

A 40 MVA, 132/33 kV transformer was transported to the site with oil filling. After installation, the transformer was topped up with oil transported in drums. Drums were lying outdoors for many months before the oil was taken out.

On the pre-commissioning check, it was found that core-to-earth insulation was almost zero. Based on urgency, the user energised the transformer against caution from OEM and kept it under no load for a day. When the reason for the poor insulation resistance of the core to the ground was investigated, it was found free water was remaining at the tank bottom for 5 ~ 10 mm level, earthing the core lamination edge to the tank.

Some quantity of oil along with water was removed from the tank bottom sludge valve and collected in an oil tank. This oil was then dried out separately by hot oil circulation. Suspecting wetting of paper insulation, transformer insulation was dried out by hot oil circulation (HOC) – dry nitrogen filling – stabilization for 12 ~ 24 hours – vacuum keeping at less than 1 mm level for 12 hours - hot oil filling - HOC cycles.

The water must have entered the transformer from the oil drums. When free water enters the oil drum, water being heavier, separates at the bottom with oil at the top. When oil is sampled out for testing, only top oil will be sampled. To avoid this, oil in the drums shall be sampled using a sampling thief (pipette-like, long glass tube which will reach the bottom of the drum, or a water-seeking paste may be used for detecting free water in the drum.

(2) Case history: 2

A 150 MVA, 220/132 kV class Auto-transformer tripped out immediately on the first energization. The winding arrangement was tertiary-common-regulating and series, in that order from the core.

 A flashover on the regulating winding at the top lead take-out bunch was noticed during internal inspection.

Moisture might have entered during transport/installation (the bushing turret was exactly above the failure point which might have been kept open too long) and water collected in the thin paper insulation covering the lead takeout. A more probable fiasco might be a sudden drizzle during bushing installation and raindrops falling over the lead insulation and causing localised severe wetting.  Localized wetting is hardly dried during hot oil circulation and may not influence the overall moisture in oil measurement by the Karl Fisher method. Shorting of tap winding, consequent to flashover from the start of regulating lead to finishing end caused heavy current flow in the regulating winding, many times more than the usual short circuit current flow for which it is designed. Still, the regulating winding withstood the heavy fault current, thanks to the robust design and winding construction using an epoxy-coated continuously transposed cable. The transformer had to be brought back to the factory for detailed inspection, repair, drying out and retesting.

10.  Bibliography

2008- CIGRE Technical Brochure 349-Moisture Equilibrium and moisture migration within Transformer Insulation systems, WG A2.30- Pages

2011-  CIGRE Brochure 445 -Transformer Maintenance Guide -WG A2.42 - Pages 122

2015- IEEE Standard C57.94- Installation, Application, Operation and Maintenance of dry type Transformers

2019- IEEE Standard C57.93- Installation, Application, Operation and Maintenance of oil filled Transformers

2021- CIGRE Technical Brochure 857- On-site Assembly, on-site building and testing of   Transformers