Saturday, August 4, 2012

Could more intelligent computers have prevented the ever biggest power outage in India?

Yes – and No! It all depends. Computers do what we want them to do. They don’t get tired, work 24*7, are reliably doing their job. They do what it has been told by a specific program and configuration. And then there are a lot more crucial aspects to take into account.

What is needed are “intelligent actions to correct problems” in due time. The intelligence may be implemented by “smart” humans or “smart” computer programs. More important: Very crucial requirements for a stable system are the various settings, build-in redundancies and the various reserves (generators, lines, head-room in power flow, transformers, …). These requirements are specified by humans! Depending on the level of risk responsible people are willing to accept, these requirements may vary greatly from one utility to another.

Depending on the settings used, build-in redundancies and the amount of the various reserves (generators, lines, transformers, …) the total system costs may be low, moderate or high! Reducing redundancies makes the system less stable – in general. Reserves and redundancies could be quite expensive.

Building a stable system is not problem in principle – it could be build if you have “unlimited” resources. The question is more: What is the maximum cost an utility is willing to spent to meet a certain risk level? Risk analysis and the level of risk accepted are key – and how and when the operators use the reserves. If an operator uses a reserve in normal operation, he cannot use that reserve in a critical situation again – you cannot eat the cake and have it.

The power delivery systems are very complex – most people do not care what it means to plan, design, operate, maintain, extend, and use such complex systems! Computers, high speed communication and even IEC 61850 are all just tools. Even a fool with a tool is a fool. And: A fool with a tool can foul up a system much faster than a fool without a tool.

The most crucial influence on power delivery systems is man-made! During a seminar an electrical engineer told me that they had a lot of serious discussions with the accountants and management on how many transformers they were allowed to replace per year. They agreed to replace two per year. Great! But: They utility had 300 (!) transformers in operation. That means: It wiould take 150 years to finish the replacement program! Unless …

Any question?

No comments: