Establish Design Rules and Build the Grid

I will start with the design of the grid. In this section I will be presenting the following topics:

• • basic modeling and designing of grid elements
• • establishing rules for acceptable levels of resilience
• • some examples of acceptance criteria
• • how a decision on acceptance criteria can affect building of infrastructure
• • assessing stability with an example of an actual event
• • designing to survive a solar magnetic disturbance

As mentioned in chapter 1, there are engineers who plan the development of the grid. This is done in long-term studies, usually at least ten to fifteen years out. Their inputs include all new transmission and generation construction projects presently in the plan, load growth estimates along with estimated locations for the growth, new generation plans, any generators (or major loads) that may retire, and any consistent issues that have come up in the operation of the system. They also use similar forecast information from neighboring utilities.

What the engineers do with all this information is they build a huge model of the electric system that they can use to test multiple scenarios. The scenarios tested include a number of outages of major transmission lines and/or generators in various combinations. What they are testing to determine is if the flow of electricity and the voltages available at critical points on the grid for these conditions are within a defined range.

There are rules that the industry folks have developed over their years of doing this that define the acceptable range of results. These rules were developed under the auspices of an industry group called NERC (North American Electric Reliability Council [now it is a corporation]). NERC was formed after the northeast blackout of 1965, when the federal regulators were threatening to take a more active role in the operations of the grid. The industry formed NERC to demonstrate that they could monitor and control themselves using peer pressure. Standards were developed by experts who worked in the industry, and all players vowed to comply. The point I want to make here is that there are standards that the industry follows that describe the limits that the grid must be designed to meet. But consider this: as the criteria for design include increasingly severe situations, utilities must build more infrastructure.

I will present some examples of planning issues to help explain the kinds of decisions that are made. The easiest-to-understand example is the rule that the system must be designed in such a way that the loss of any single element (e.g., a transmission line or a single generator) does not cause the loss of grid stability, nor does it require any manual load shedding to avoid overloading other transmission elements. The golden rule for planning and operating the grid is this: the system must be able to survive a single loss of an element of the grid. This is called a single contingency.

When a networked transmission line is lost (tripped), the current that was on that line is instantly transferred to other, available paths. This should imply that designing a system that relies on transmission lines being loaded above 50 percent of their capability is risky, since the loss of nearby lines would result in an immediate increase of the current on the line. As a footnote to this design concept, I’ll mention that in the early nineties, the opponents to building transmission lines said something like, “Their average loading today on their transmission lines is below 50 percent, which proves that the lines are already overbuilt.”

I would like to provide a little additional detail on this single-contingency concept. Consider as an example a situation where studies show that a generator is needed for support in the case where a single transmission line has tripped off-line. Now the question that must be answered is if there is concern about the loss of the line when the generator is not available. Generator availability in the 90 percent range would be considered pretty reliable, but that would leave 10 percent of the hours (876 hours per year) without the local support of the generator. Would the rules allow the grid to be at risk for that many hours every year? Will the rules allow for operator intervention in this case?

The simultaneous loss of a transmission line and a generator can be a very severe contingency for some systems, especially when the reactive component of electricity is considered. Operator intervention, which would include something like redispatching generation, or shutting off large segments of load, may be necessary to survive this under peak loading conditions. The decision of which way to go on this point can be a daunting one to make, as the decision to design the grid so that this contingency does not require load reduction may result in the need for a new transmission line. The outcome may be that the simultaneous loss of generator and line requires load shedding, which could go on for days, weeks, or even months if the line is severely damaged and the generator is not available. And to make matters worse, the load shedding could be required over a large area of load, such as a large city.

Of course, there are many potentially worse situations than the simultaneous loss of a generator and a transmission line (for example, the simultaneous loss of all the transmission lines on a right-of-way) where the rules would say that a load shed is acceptable, but the grid must be able to survive. Decisions made that determine the extent of operator intervention assumed in these cases become critical to the amount of infrastructure needed. That is, if the assumption is made that the grid must survive without any operator intervention, then there will be a need for additional lines or other elements.

Planners also look closely at the possibility of an unstable generator for various conditions on the grid. Previously I mentioned what happens when an electrical perturbation of some kind (for example, a short-circuit on a nearby transmission line) occurs and the generator “wobbles” a bit with the grid. This happens regularly, and the planners must ensure for the operators that the wobble is damped. If the wobble isn’t damped, the generator will normally be tripped off-line, hopefully before there is some damage to it.

The electric company for which I worked had an event in the middle of the night sometime in the early 1990s that I still shudder to think about. A large transmission line suddenly tripped off for no apparent reason, and this created instability between our pumped storage units, which were pumping water at the time, and other units on our system. The wobble I was talking about earlier didn’t dampen; it got bigger and bigger! Fortunately, our shift supervisor on duty immediately realized what was going on and ordered the pumping units to be tripped off-line one at a time. There were five units pumping at about 400 Mw each, to put some perspective on the size of this event. After the fourth pump was tripped, the system calmed down.

This perturbation was felt throughout the eastern grid. Had our planners missed something? I don’t think so. You simply can’t study every possible configuration of the grid. The good news is that after this scare, automatic controls were put into place on the pumps that prevented this instability.

I mentioned solar magnetic disturbances (SMDs) earlier. There are some basic things that the planners can do to help the grid survive these events. One thing that has helped is that resisters were installed into the path along which the SMD-caused ground current would flow. This is done where the transformers are grounded. By doing this, the current flow from the earth’s crust through the grid is reduced. Another thing that should be done is to make sure the protective relays are designed to take into account the possibility of SMDs. Does doing these things prevent the next SMD from causing a grid collapse? Since we have no idea how intense one of these events can be, there is really no way to know the answer.

I think the grid can survive with little trouble the highest-magnitude SMD that has been seen since the grid was developed. Also, if an extreme event occurred and managed to cause a collapse, the real questions relate to how much damage has been done and how quickly grid operators can get the grid back up. My guess is that the damage would be minor and localized, and that the grid could be completely back in a few days at worst. There is no reason to lose sleep over an SMD-caused grid collapse.