This is in relation to the article by Dr. Chandre Dharmawardana, “Sri Lanka’s power supply, blackouts, and how to prevent them” published in the Island Newspaper.

I am an electrical engineer, who has spent the last 30 years in research and development in telecommunication systems and system reliability, which includes software reliability. Today, as a consultant, I am guiding the engineers and scientists of the Canadian Utilities in trying to understand a previously unknown condition affecting the grid networks. Canadian utility networks are highly sophisticated, complex networks, protected by many strategic and tactical networking designs, to account for not only the inevitable faults, but also the ability to withstand weather from -50C to +50 C, including rain, ice and snow storms.

Over the last five to six years, the utilities have found the prevalence of large magnitude, 7th and 9th harmonics, which has been contributing to several unforseen outages in the Canadian networks. This phenomenon has been observed in Europe and in the USA. The grid is primarily designed for large 3rd harmonic fault tolerance. Experiments and tests over the last five years seem to suggest that the switching power supplies, wind and solar power generation equipment, and the large plethora of small mobile telephone and laptop charging equipment might be the source of these large magnitude harmonics. Yet proper fault diagnosis and containment had prevented outright failures, even in the presence of fault conditions previously not encountered or understood. That is good system effectiveness.

The point however, is that rapid problem sectionalization, fault containment (from rapid fault propagation), rapid problem restoration and preventive maintenance are at the heart of designing and maintaining mission critical systems, such as the utility network in Sri Lanka. The system complexity is needed in order to ensure rapid problem sectionalization and containment. As Dr. Dharmawardana suggests, it is inconceivable that the actions of an individual of the CEB is to be blamed for the catastrophic outages the country had experienced. The competency of the individual engineers at the CEB must be just as good as the competency of the medical, public health, the military and the police personnel, who were primarily responsible for containing and controlling the transmission of the Covid virus in Sri Lanka; comparatively better way than most other countries in the world. Having worked in the UK, Canada and in Europe, it is my informed opinion that the engineers at the CEB are no less competent than any other engineer in the world.

Hence, the engineers in Sri Lanka must be competent in performing Failure Mode Criticality Analysis (FMECA) in the mission critical systems such as the electricity grid network. It is so fundamental, that any graduate engineer is taught that in mission critical systems, one must account for all faults down to at least the third degree of failures, and design the system to be fault tolerant to that level. There are no such systems as ‘fail safe’ systems; failures are a part of any complex system, which includes all human induced failures.

The system should be designed to be robust enough to withstand at least three levels of faults before catastrophic failures precipitate. Telephonic systems are designed for no more than one catastrophic failure in 60 years. However, if the operators have no discipline, and are so incompetent, that they allow the faults to propagate to the second, third and the fourth levels, without containment, the catastrophic outages are inevitable. It is really “stupid’ to suggest that just one operator pulling one circuit breaker brought down the entire network. This is inconceivable in a mission critical system. As we have seen, Boeing and the FCC, not performing the FMECA diligently and independently, partly accounted for the recent fatal aircraft accidents. Testing software systems against requirements, either specified, or deduced from FMECA, is an absolute necessity in mission critical systems. In order to effect cost savings, sometimes the systems are only tested against the specified requirements, and not against the deduced requirements from failure analysis. This is not the norm, but it happens elsewhere in the world.

In the case of the CEB, the only conceivable possibility is total mismanagement. Lack of effective preventive maintenance, systemic or rampant organization issues that prevent the effective utilization of the collective knowledge base that exists within the CEB; these are refactors at the heart of these outages. As a trained engineer, it is my opinion that it is the lack of process and operational integrity that caused the outages. These elemental issues must be addressed as a matter of priority, in order to eliminate such outages in the future. No amount of artificial intelligence and smart system designs can eliminate the need for effective and smart operational practices.

LAYANANDA ALLES, Eng.Montreal, Canada