The recent explosion at Biyagama sub-power station which led to a countrywide power outage occurred due to negligence, lack of maintenance and inspection by the Ceylon Electricity Board (CEB), a technical report developed by a panel of experts points out.
An expert told The Sunday Leader that CEB engineers should have been more cautious in this regard as it directly comes under their purview. He said that in the past, engineers were very concerned about their responsibilities and conducted regular checks. The report which is with the CEB’s top management indicates that a comprehensive study is of paramount interest and a necessity to avoid such occurrences in the future.
According to the report, the failure of transformer No.2 had no effect on the restoration of the transmission system. However, the restoration process was unsuccessful four times. Hence, a comprehensive study is necessary to avoid such occurrences in future. The Biyagama Receiving Station (BRS) was constructed to receive Mahaweli power generation in the early 1980s. The station receives power at 220 kV and switches to Kotugoda, Pannipitiya and Gas Insulated Switchgear (GIS) Kelanitissa. Also, the 220 kV power is stepped down to 132 kVand 33 kV voltages using 2 Nos. of 250 MVA, 220 kV/132 kV/33 kV transformers. At 132 voltages, power is distributed to Sapugaskanda Grid Substation and receives power from Sapugaskanda Power Station. In addition, there are 8 Nos. of 33 kV feeders available to distribute power.
A detailed technical report says that the load tap changer of transformer No 2 exploded due to a fault in the diverter switch.
“Consequently, the lines connected to bus bar No.1 tripped off resulting in the failure of the total transmission system on March 13,” the report said. “It also revealed that the failure of the diverter switch was due to lack of inspection and maintenance of tap changing equipment.” Therefore, the report recommends that inspections should be carried out on all sister tap changing equipment at regular intervals as stated in the relevant operating instructions of the manufacturer. The report also highlights the evident misses in routine maintenance work and says, “This is an alarming situation and prompt action should have been taken to test dielectric strength of all remaining sister tap changers of BRS and Kotugoda Grid Substations. But maintenance records of BRS do not show such testing had been done.” Operation records of the day the incident occurred shows that the restoration process had been unsuccessful four times. The report which recommends immediate action to prevent similar failures in the future said inspections should be carried out at regular intervals as stated in the manufacturer’s operating instructions. “Load Tap Changers (OLTC) are a complicated device and as such inspections should be carried out by highly trained personnel.”
When contacted, Ceylon Electricity Board Engineers Union (CEBEU) President Athula Wanniarachchi says the explosion of tap-changers of high voltage transformers has been investigated by internal and external committees.
“Our members extended their fullest cooperation for these investigations. As far as we know their observations and recommendations have been handed over to the higher management of CEB. We are waiting for the report. We believe the investigations had been conducted in a professional manner,” he stressed. “This is an isolated incident and it should not be generalised as to say engineers are not maintaining the equipment properly. If not for proper maintenance, can equipment that is over 30 years old last this long? Meanwhile some equipment manufacturers are trying to “fish in trouble waters”, he added. Wanniarachchi alleged that some manufacturers were giving false information even to the Prime Minster and ministers that their proposals for major overhauls spending millions of rupees were not implemented by CEB engineers stating lack of funds.
“It is very easy for our engineers to replace all major assets every 10-15 years. But that is like investing on an entire power system once every 20 years.
Can any government afford to invest only on improving the quality of electricity? The CEB must be appreciated for maintaining these assets this long. What these unwarranted criticisms would ultimately do is, make all engineers “play safe” and recommend replacing assets with new ones instead of maintaining them and taking a risk,” he said. CEB in the right direction CEB Chairman Anura Wijayapala told The Sunday Leader that with reports and recommendations they are now on the right track to mitigate and rectify issues. He said it is the duty of the engineers to find solutions to providing the best service –uninterrupted power supply at the cheapest cost for consumers.
Even Power and Renewable Energy Minister Ranjith Siyambalapitiya strongly believes the CEB Chief is the right person for the job and for the betterment of the CEB. The Sunday Leader learns that Japanese experts had earlier given CEB engineers a programme for transformer maintenance which the engineers had followed. But the German experts say that the maintenance programme was wrong and that the moisture in tap changers was too high. They recommend changing oil in tap changers. The transformers had caught fire since the tap changers exploded. German experts have still not given a final conclusion. A number of committees showed the reasons and made recommendations. A number of energetic young engineers who have experience in generation, transmission and other sectionswere appointed to the committees. They are Sreepal Gunaratne, Janaka Aluthge, J. Nandakumar, Kusum Shanthi, Dr. H.M. Wijekoon and Lakshitha Weerasinghe. They were appointed under the approval of CEB Chairman and General Manager M.C. Wickramasekara. The team had been given a free hand to mediate and propose short-term solutions to come out of this crisis. The Cabinet subcommittee appointed by the President held discussions five times and proposed mid-term and long-term solutions based on those recommendations. A mid-term solution was that big firms like the Harbour and Ceylon Petroleum Corporation have generators of the capacity of 70 MW. CEB proposed that they generate their own power and that CEB would bear the loss incurred by them.
“We are discussing the long-term solutions. We should hand over that report to the President shortly,” a senior Ministry official said.
The dynamic response of the system following the event was captured in the transient recording provided to MHI (Manitoba Hydro International) by CEB.
Upon the tripping of the generating units, the system frequency dropped and as expected, UF load shedding kicked in, in an effort to stabilise the frequency. The system was operating at a light load prior to the event. Thus,the system was vulnerable to over-voltage situations. Load shedding and transformer on load tap changer actions aggravates this condition. This is because the net capacitive reactive power generated by the lines increased in response to the reduced loading.
As the system voltage goes up, the transformer on load tap changers (230/132 kV and 132/33 kV step down transformers) act to regulate the low side voltage (LV side). A result of this action is an increased level of reactive power injection to the HV side of the system. As a result, the above system voltage in the high voltage network reached levels that activated further tripping of equipment and eventual system collapse. The system collapse of October 2015 was a ‘slow’ event lasting nearly three minutes. This is a further indication of the tap changer action (a slow action) on the overall system response during this event. The system model of the Sri Lanka power system was provided to MHI by the CEB. MHI performed preliminary load flow studies (steady state) and had a number of discussions with CEB engineers to gain a good understanding of the system characteristics and the specific event background.
The discussions also identified potential limitations of the System Dynamic models provided to MHI in PSS/E format. The simulation results based on the model provided by the CEB were compared with the actual event recordings. The system response predicted by the model is far removed from what was observed during the event. Specifically, the model predicts an extremely stable voltage response following the event. Thus the model had to be adjusted. The ‘adjusted’ model followed key trends of voltage variations observed during the event. Two key adjustments were made to the model based on MHI’s past experience on similar situations. The tap action of transformers can have a significant impact in a ‘slow’ (also referred to as long-term) voltage instability situation. The model was updated by implementing the tap changer actions based on information from CEB. Reactive power absorption limits of generators were adjusted. The recordings clearly indicated the gradual, uncontrolled voltage increase.
This can only happen when generators (the only dynamic reactive power regulating device in this system) reach their limits (exciters reaching their lower limit). The model adjustment was a necessary part of the overall study. The important consideration was that the model showed the same trends as those recorded following the generator tripping.
This ‘adjusted’ model was used to perform further dynamic studies to make overall conclusions and recommendation.
The following are the main observations and conclusions of the study. ‘Based on MHI’s extensive experience on similar situations, it is our opinion that the ‘adjusted’ model sufficiently represents the Sri Lankan system for the purpose of examining the event of September 2015. However, the ‘adjusted’ model should not be used for general purpose system planning studies.
MHI strongly recommends the following: a. the model dynamic data be reviewed and updated based on accurate name plate information, test results and data gathered through site visits. b. The existing dynamic model is not suitable for long-term event analysis such as the event under consideration.
Proper model update including transformer tap changers and specific load modelling is recommended (i.e. induction motor models and composite load models representing actual load behaviour should be used instead of simple static load model used in existing models). Based on generator ratings provided to MHI (steady state limits provided in PSSE data), it is very likely that the full reactive power capabilities of generators are not being used. The generator reactive power limits as well as the actual settings that are currently in place should be readily obtained from the respective generating stations.
If the reactive power limits are not at their respective maximum values, this should be reviewed and adjusted where possible. As per the load flow data provided to MHI, the system steady state voltages on many 230 kV and 132 kV buses were above 105% (1.05pu). The international practice is to maintain the steady state voltage on the High Voltage network at or below 105%. Allowing 110% voltage at steady state will bring the system devices dangerously close to their over-voltage protection limits.
It is recommended to maintain the system steady state voltage within ± 5%. The present CEB criteria is ±10%. This can be achieved by mechanically switched shunt devices (reactors and capacitors). Simulation verified that this is a feasible solution to prevent system instability under similar events in future.
The size and location of such shunt devices should be determined through system studies once the model is updated and validated.
As an interim measure, the steady state voltage during low load periods can be improved by ‘tap staggering’ of parallel transformers (though not recommended as a permanent solution). Taking selected transmission lines out of service during low load periods (though not recommended as a permanent solution as this action impacts overall system reliability). Shunt reactors are an effective means of controlling steady state voltage as well as improving the system dynamic response. The effectiveness of shunt reactor additions is demonstrated through simulations.’
The Sunday Leader investigations reveal that blackouts were frequent since 2009 due to transmission and voltage levels. However, surprisingly, nobody seemed to be concerned enough to do something in this regard.
According to Kanthar Balanathan, former Director and Specialist Power Systems Engineer (Australia), going by the media reports the German specialists, after investigations, revealed the explosion was in the diverter switch, which is the On Load Tap Changer (OLTC). The Load Throw Off (LTO) of the capacity of this Biyagama transformer also causes stability issues and tripped the generators at the Lakvijaya (Norochcholai, Puttalam) power plant. The capacity of the generators at Lakvijaya Power Station (LPS) is 3 x 300MW. The Transformer at Biyagama is rated for 220/132/33kV.
According to the report the feeder lines tripped after the transformer caught fire at Biyagama.
Balanathan says that the statement cannot be true. It may be the first incident of a trip. During the period when this disaster occurred at about 2.15 p.m. there was no lightning or rain. The environment too, was normal. Indeed, there was no rain at all in the region. Another transformer explosion at Kotugoda caught fire similar to the explosion at the Biyagama grid sub-station.
Both transformers are similar single phase auto-transformers. It could be a “Type Fault” on these types of transformers. Type: Takaoka Autotransformer (1983) 132//33kV (secondary), 200/3, 250/3, 60/3 MVA, YNa0d1, ONAN/ONAF with 13 taps on both. There are three of this type one on each of the phases.