|
|
|
Brainstorms
Electric Grids
by Stephen W. PhilbrickEveryone knows about the Northeast blackout last August. Not as many know about the Memphis area blackout in July, even though it affected over one million people for up to three weeks. The September black-out in Sweden and Denmark affected five million, but also garnered limited press. A smaller blackout in London affected half a million people in August, and over 50 million in Italy were affected in September.
What, you may be asking, does this have to do with actuaries? I didn't see an immediate connection either, until I read a fascinating blog entry from Steven den Beste (http://denbeste.nu/cd_log_entries/2003/08/Un-blackingout.shtml) explaining why it takes so much time to bring a grid back online. Exploring some of the links, I ran across an entry by Rich Hailey (http://shotsacrossthebow.com/archives/001778.html#001778) with a basic primer on how electrical grids came to exist.
In short, Hailey explained that a local system needs to have more generating capacity than its average needs, to accommodate peak needs. But this excess capacity is idle most of the time. Excess generating capacity costs money, so utilities would like to minimize it, while customers (who dislike power disruption) would like as much as possible to maximize it. Neighboring utilities realized that, to the extent peaks didn't exactly coincide, they could jointly get by with less than the sum of the required individual excess capacity if they hooked their local systems together. As more local systems were connected, the ratio of excess capacity to average capacity dropped, increasing the efficiency of the overall system.
It struck me that the excess capacity of a local electrical system could be analogized to the need for surplus in a stand-alone insurance company. The company needs to carry expected reserves (comparable to average electrical loads) plus an additional amount in surplus (comparable to excess capacity). Customers want the surplus levels strong because they want the company to pay all its claims, but the owners don't want too much surplus because it reduces the return on surplus.
Connecting local electrical grids is analogous to reinsurance; not in the sense of stand-alone reinsurance, but quota share reinsurance between two primary carriers.
The analogy is less than perfect of course, but the question in my mind is whether the analogy is good enough to gain insights. Is the science of electrical grid connections advanced enough that the insurance industry could learn from it? Conversely, is the insurance/reinsurance model advanced enough that the electric utilities could improve their operations by borrowing ideas from us? Or maybe both?
One of the criticisms of the electrical grid model is that no one seems to be responsible for the interconnected part. Their "reinsurance" is solely analogous to primary companies engaging in mutual quota share. There doesn't seem to be the analogy of a professional stand-alone reinsurer. Should there be?
I'm guessing Utility A doesn't get unrestricted right to draw upon the capacity of Utility B. For example, if Utility B is also experiencing a peak, it needs its capacity. Perhaps then both utilities would draw on Utility C. If we designed a reinsurance contract in an analogous way, we might include several parties with clauses indicating that participants with higher loss ratios bear a smaller portion of the shared burden. This obviously has potential for problems, although some can be mitigated with industry loss warrantees (clauses which force coverage to apply only if industry losses exceed a certain level).
These are only two examples, based upon my very limited knowledge of the electrical grid system. However, I have to wonder if an electrical engineer designing the electrical grid and an actuary thinking about improving the capital structure of a company might find they had a fair amount in common, if they avoid the technical jargon and talk about the overall structure of the respective industries.
