As in zero, nada, zip, zilch, nothing.
Do you think that the very high level, high profile U.S-Canada power system outage task force could have spent all that time and money investigating the blackout, and then come to the recommendation that nothing needed to change? No way.
But on the face of it, why not ?no change?? We may not need more reliability, or at least not more of a mandatory, uniform level of reliability. As the blackout report points out, big blackouts are not very common. The electric power system is already highly reliable, but every ten years or so a significant blackout comes along. Seems to be a pretty good record.
Big blackouts are costly, and ?never again? is a fine sentiment, but do we really want more reliability? Consumers already pay a lot for reliability, and whether we want to buy more depends upon how much will it cost us. The task force report has given us a long shopping list of top-down recommendations, but it didn?t include a price tag.
There is an even deeper problem. Does anyone have a good idea whether the typical electricity consumer in the United States or Canada would rather have more reliability at higher power prices or less reliability and lower power prices? Some consumers would probably be willing to pay more to have more reliable service, and others would pick lower prices even if service quality went down a little. The existing top-down system offers no way for operators and policymakers to gather information on how different customers value reliability. Nor does it offer much in the way of buying different levels of reliability, instead treating reliability as a ?one size fits all? characteristic. Note how this approach reflects the supply focus we discussed on Friday.
Both of us have in effect chosen some of what reliability choice is available, having purchased battery-backup power supplies to keep our computers up and running even if the local distribution company is having problems: paying a little more to have more reliable service for a select appliance or two, and needing a little less average reliability from the local wires company. Businesses do the same thing on a larger scale, with companies that have special needs for highly reliable electric power spending millions of dollars to secure their supplies. This is a focused demand side approach that provides very targeted system reliability, which we?ll discuss on Tuesday.
What ability do consumers have to get the qualities of supply-side power system reliability for which they are willing to pay? Are there other ways to provide reliability that don?t have the pitfalls of the uniform top-down system?
Hung-po Chao and Robert Wilson suggested one approach, called ?Priority Insurance,? in a 1987 American Economic Review article. The essence of the idea is to have the electric company pay consumers when the lights go out. A simple idea, but they add a twist: the electric company offers different qualities of service. For a higher price, you get a lower probability of being cut off when the system is short of power (and a higher payment from the electric company when the lights go out); pay a lower price, get a higher probability of being cut off (and a lower payment). Customers would actually be able to choose between price and reliability, based on the individual tradeoff they perceive between them.
Charles Nossair and Dave Porter tested a version of this idea against a simple system of proportional rationing of shortages, and found that their version of Priority Insurance was more efficient. Nossair and Porter added another twist in that in their version the number of levels of service and price levels were endogenous to the customer evaluations ? it sounds more complicated, but it is easier than having the electric company try to figure out the ideal set of offerings, and more adaptable, too. (1992, Journal of Economic Behavior and Organization)
While one benefit of Priority Insurance approach is that it allows the electric company to allocate a shortage efficiently by having customers prioritize their own use, a bigger payoff comes from the information created through consumer actions. The Chao and Wilson approach enables consumers to express values separately for energy and for service reliability, allowing the company to target investments where the can provide the most value to consumers. Companies could distinguish what customers are willing to pay for power from what they are willing to pay for reliability.
We already have a pretty reliable grid ? maybe even too reliable in that some consumers would rather have lower prices and a little less reliability. How much more do consumers need to spend? The answer may be a big, fat zero.
What we need zero of is top down change, and what we need more of is bottom up consumer choices producing information and incentives to make the system better able to provide the diversity of reliability that consumers need. More bottom-up consumer-driven influence over the provision of system reliability would make the electric power system move toward a dynamic, robust, self-adapting organic network that is the most reliable system humanly possible.
If you think that ?no change? had better than the proverbial snowball?s chance in the media spotlight, then you don?t know nothing about politics, regulatory policy, or very high level, high profile international task forces.
It is in the nature of the process that something had to be recommended. Yet sometimes wisdom lies in know what not to say. It would have been a very wise 0.
“what we need more of is bottom up consumer choices producing information and incentives to make the system better able to provide the diversity of reliability that consumers need.”
Of course the information and incentives would have to be very reliable and involve little additional cost … Then we could also have the reliability and accuracy of this information and incentives subject to a sliding scale of costs for how much reliability consumers would be willing to pay for … and then we could outsource …
I do have a couple of problems with the “reliability as a commodity” concept.First, everyone depends on services that depend on reliable power supply; this places at most one degree of seperation between the individual consumer and the need for reliability. For example, telephone providers, cell providers, water pumping and treatment, sewage processing, emergency services (police, fire, EMS, 911), traffic signals, hospitals, and food suppliers (think frozen food storage and distribution) all provide basic services that hold our society together. Without reliable power, we could be left without critical services. Keeping these alive helps everyone on a day-to-day basis and in emergency situations.
Second, reliability is not a point-and-shoot “good” that can be delivered to one place and not to the surrounding areas, nor will that possible in the foreseeable future. You can’t offer 99% reliability to a load (customer) if the entire chain from generation to transmission to distribution doesn’t have at least 99% reliability; and if the grid does have that level of stability, then everyone connected would automatically have 99% reliability so its additional value would be zero.
Ben,
As Lynne pointed out in the posting, those who require higher levels of reliability have alternatives available to provide the required reliability. For example, the FAA regional air traffic control centers have massive UPS systems consisting of battery banks and engine-driven generators, to assure that air traffic control can continue during even very long power outages. Bank credit card processors have perhaps the most sophisticated emergency power systems currently in private service, because of the high costs of service interruption. Telephone service is ususally available during power failures, although individual electronic phones may not function. Most businesses with critical loads have emergency power systems to support those loads.
Consumers who have not installed emergency power systems have made a decision (conscious or otherwise) that the cost of increased reliability is higher than they are willing to pay. Long outages caused by major failures or extreme weather frequently cause these decisions to be revisited, as has been the case following the Northeast blackout.
The concepts of reliability insurance and reliability contracting are not new, although they are not common in the utility industry. Forty years ago, my employer had a reliability contract with a shipper in a situation which required “just-in-time” inventory delivery. If the shipper failed to deliver on time and the production line was shut down, the shipper was responsible for the cost of the lost production. The customer and the shipper negotiated a contract which recognized both the extreme cost of 100% reliability and the cost of lost production. The shipping costs for other shipments to the same customer and the shipping costs to other customers with lower reliability requirements were unaffected by this arrangement.
It is important to recognize that not all potential solutions to specific customers’ reliability requirements also require that all other customers’ reliability levels increase as well. The technology exists to provide very localized and even customer specific solutions. With this technology, the “grand solution” may well not be the best solution.
“Priority insurance” is a good idea, though it misses part of the problem, which is that system-level remedies are only likely to work for system-level outages, while most outages are local. Ben Minshall’s comment hints at this issue.
Many power outages (in my area, almost all of them) are narrowly local, typically the result of lightning-induced damage. The power company in such cases would be unable easily to allocate replacement power, even if affected customers were willing to pay higher rates to be eligible to get it, because the nature of the problem is that last-leg transmission lines into people’s houses and businesses are blocked. Probably many of us buy power backups for our computers to handle short-term problems like these. Self-insurance is cheap for such cases.
To deal with the more general question of whether the power grid is not reliable enough, reliable enough, or too reliable, it might be more useful to try to estimate how many home and business owners buy backup generators. Generator purchase statistics might be a better proxy for the demand for insurance against all kinds of power outages than demand for priority insurance (which would protect only against high-level grid failures) would be. The fact that few of my acquaintances have bought generators suggests to me that the power grid is more reliable than most customers need it to be.
Perhaps the best approach would be for power companies to offer combinations of priority insurance (lower rates for power customers who need less system-wide reliability) and real insurance (cash payments to cover lost productivity for customers who want protection against all kinds of power outages).
This is certainly an interesting issue. Thanks for raising so many good questions.
I wrote my comments before reading those of Ed Reid. The meta-question is how most economically to insure against power outages. As Ed points out, big business has already worked out answers for itself. It will be interesting to see how much demand there is for answers on the small-business and individual levels.
Ed,
When reading my post again, I notice I never got around to mentioning extended outages as a possibility. We got lucky in the August blackout in that not much generation and transmission equipment was damaged. However, if there was significant damage to generators, certain stations could be offline for weeks during the repairs. There are areas of this country where having one or two stations offline would require rolling blackouts for the duration of the repairs. These are the type of reliability issues that need to be addressed on a system-level. Well placed terror actions could have similar effect, and we qutie simply aren’t capable of dealing with that right now.
Regarding UPSs and backup generators, I agree they are essential for managing local-level reliability, but most sites only have enough diesel on hand for a few days at most. Since most power outages are the result of lightning, trees, or car-pole accidents, only a few hours is expected. Although I do see your point that an organization that needed 6 weeks of uptime could make a serious diesel fuel investment.
Jonathan,
The individual small customer perception of the value of insurance against power outages is very different the day before the failure and the day after, because most small customers do not plan for power outages. On the day before the failure, or the day before the hurricane warning, dealers have ample supplies of emergency generators. On the day after, there is not a generator to be found anywhere at any price. Businesses plan ahead and do cost / benefit tradeoffs.
I have kidded for years that if natural gas utilities had begun by installing pipes on poles, all utilities would have been underground a long time ago. If you catalogue all of the reasons for grid failures, the listing gets pretty thin beyond damage caused by uninsulated line contact with trees, falling trees, high winds, ice storms, hurricanes, tornados, lightning strikes and automobile accidents. Murphy’s law effects can be minimized by conservative design, but they can’t be avoided. Planting everything underground doesn’t avoid all interruptions either – the primary cause of natural gas service interruptions is the construction backhoe!
Ben,
FAA uses natural gas fueled engines, as do many other standby / emergency power users.
Reportedly, one of the major reasons the outage extended as far and as fast as it did is that the safety systems employed at power generators acted faster than the safeties on the grid. This was not an accident!