Smith/Kiesling Wsj

Vernon Smith and I have a commentary in today’s Wall Street Journal (subscription required). Entitled “Socket to Arnold”, our point is that the existing system of average pricing is unfair to off-peak users. Off-peak users subsidize peak energy use, and that’s both inefficient and unfair:

But current policy unfairly forces consumers to pay rates based on the average hourly cost of energy and industry capital investment. As a result, peak utility cost is much higher than what consumers pay, and off-peak and weekend cost is much lower than what consumers pay. The utility earns an abnormally high profit from off-peak consumption and loses money from peak sales. Peak period sales are thus subsidized by the implicit transfer of funds from the utility’s high profit on off-peak users. In effect, utilities profit on energy needed to dry clothes at 6 a.m. and subsidize clothes dried at 4 p.m.

Relieving this disparity requires a significant shift in thinking and corrections in the regulatory environment. Under the current regime, local utilities enjoy a government-protected monopoly on energy and the wires for delivering it — bundling the sale of energy with an additional charge for the rental of the facilities and wires leading to homes and businesses. Divesting these separable activities — the market for energy and the market for delivery — would create a competitive environment that benefits customers while opening opportunities for new market entrants and the development of new technologies.

We then propose some options that would take advantage of splitting the energy market from the wires market, which makes economic sense because the energy generation portion of the industry is not subject to the same network characteristics that the wires are. So why regulate the bundle if only one part of the bundle has the traits that induce us to regulate an industry?

Unbundling the energy commodity from the wires, and from the bundled regulatory treatment, is a long-run recommendation, so we also make some recommendations for beneficial changes that Governor Schwarzenegger could implement in the short run, within the existing regulatory environment:

  • Option 3: Allow wholesale prices to be passed on to any retail customers by giving all customers a choice between a fixed average price or one of the many time-of-use pricing program technologies, such as remote appliance switches, time-of-use metering or load management systems.
  • Option 4: Allow customers to supply their own energy through a distributed generation source without charges for the utilities’ wires and infrastructure costs. If a customer uses grid wires only some of the time, the cost should be prorated hourly, giving the customer credit for hours off the grid. This requires utilities to meet the opportunity costs created by new technologies, and it removes the utilities’ ability to use regulatory cost-averaging rules to block new cost-saving sources of power. This model has the important effect of making the transmission and the distribution grid contestable, and would relieve transmission congestion.

Even within a regulated rate tariff, giving customers choice from among a menu of contracts would break the unfair and inefficient subsidy of peak use. And removing the barriers to distributed generation interconnection allows distributed power to bolster the grid, both in terms of reliability and security.

With active retail choice and potential entry of distributed generation, the likellihood of energy efficiency goes up, power plant and grid construction goes down, and it is even possible that energy use would fall. Thus markets and choice are not only good for efficiency and equity, they are good for the environment too, because demand response promotes conservation.

20 thoughts on “Smith/Kiesling Wsj

  1. The fundemental flaw in this argument is: “the energy generation portion of the industry is not subject to the same network characteristics that the wires are. ”

    Anyone interested in a more informed anaylsis should read:

    I would also suggest that there are not very many people who will do their wash at 2:00 AM to get an off peak savings of a fraction of a cent.

  2. If we were talking about a fraction of a cent, your point would be valid. There’s a big difference between the incremental cost of a kilowatt of peak versus off-peak production. It’s on the order of a factor of 2 difference in cost. It’s not unlike cell phones, where off-peak usage is essentially free.
    Having peak capacity requires plant. Actually using the plant is relatively cheap.

  3. They will if their hourly demand is several kilowatt hours, and if electricity is a substantial line item in their budgets. In other words, industrial and residential customers are more likely, ceteris paribus, to have more elastic demand. They are the ones who benefit most from active retail choice. But their choices to shift away from peak would also bring down wholesale prices for all customers, regardless of whether or not they chose to shift their demand, or whether or not they choose a price-responsive contract.

  4. Not 2 am, but people would definetly change their behavior to do washing/drying at 6am-9am, or 9 pm to 11 pm, to avoid peak our rates (1-6 pm, for example).
    Think of electrical consumption as selling tickets to a popular event. Most of the tickets go at the retail price (if you buy in advance, you might get a favorable group rate); but scalpers make a killing selling the last few tickets to people in demand. Likewise, the last 3-5% of peak electrical demand is very expensive to buy (everybody wants it at the same time!) compared to the baseline generating capacity (which runs 24/7). If incentives and technology were applied to encourage domestic and businesses to use off-peak production for some applications rather than “prime – time” power consumption, the benefits would be very attractive, indeed.

  5. We lived in Phoenix, Arizona for six years where the local power utility, APS, offered a peak/off-peak pricing plan (9:00 a.m.-9:00 p.m. M-F). We set the timer on our dishwasher to run after 9:00 each night, did laundry on the weekends, before leaving for work or after 9:00. We also bought an electronic thermostat that allowed us to set the AC at higher temperatures 9:00-9:00 and lower in the off-peak. Same with the heat in the winter (all two months of it) with higher temps in the early morning and cooler during peak. It made a BIG difference, probably at least 10 percent on an average summer bill of $200 a month and a winter bill of $120.

  6. “I would also suggest that there are not very many people who will do their wash at 2:00 AM to get an off peak savings of a fraction of a cent.”

    I recall while living in France with it’s high energy costs (and 2 tier pricing model — peak and off-peak) the washing machine we purchased has a timer on it. Load it up with clothes, put the soap and other goodies in, then when the appropriate time was reached, the washer activated. Of course the average wash cycle was over 2 hours and dryers were still largely too expensive to operate.

    If there is a measurable savings to be had and applicances had the appropriate capabilities, people will take the opportunity to use them.

    Even today in the US, there are rather sophisticated programmable electronic thermostats so the temperature can be adjusted during peak hours while no body is home. That’s a huge cost savings already waiting to be taken advantage of.

  7. At some point during 2000 wasn’t the wholesale peak-to non-peak price differential in California as high as $145/MWhr?

  8. Yes it was, so shifting away from peak would have been a very effective deterrent against generator exercise of market power.

    Interestingly, there’s a demand response pilot program here in Chicago that saw decent elasticity and shifting this summer, even though our weather was quite temperate and the ratio of peak/off-peak prices was not often more than 2.0

  9. Peak/Off Peak pricing is all good in theory, but there are a couple of significant barriers to its implementation. Only a handful of utilities have deployed electric meters in residential households that have the capability of recording Peak/Off Peak usage. Puget Sound Energy in Washington converted about 750,000 meters to an automatic meter reading system. PSE spent several hundred thousand more making changes to their billing system to handle time slice billing (as a follow up to a multi-million dollar project to write a new billing system). Then PSE went to the State to propose a Multi-Tier pricing system to the state utilities commision. The State imposed so many constraints on the experiment that it ended up doing nothing more than alienating the customer base, without making any significant changes to power use patterns. And PSE ended up scrapping the program after several months with nothing more to show for their innovative vision than a public relations debacle, and ligher wallet.

  10. Back in 1979, my parents lived in Minnesota. The local power coop had a peak/off-peak price system there, and so when the needed a new furnace they bought an off-peak heat sink. It would heat up during the night, and then use the accumulated heat to warm the house during the day.

    This is an ‘issue’ that was decided decades ago. The only reason we still hear about it is due to power monopolies and goverment management.

  11. There are several problems with the cell phone comparison. One is that electric utilities are limited on cutting off people for nonpayment it takes months usually while cell phones can be off within a couple of days of nonpayment. Two, there are no credit checks as is the case for cell phones for new customers. Utilities have to hook everyone up as long as they donít owe that company money. Three, service in outlining areas is sporadic at best. People expect electric to be provided everywhere at low cost. Four, during peak hours if someone tries to make a cell call and all the circuits are busy they get mildly irriated and try again. If the electrical grid is overloaded during peak times it does not just affect one call or customer but block and even cities of customers. See how irate customer and politicians would get about that. Oh wait, they already have.

  12. Also keep in mind that appliances could be redesigned to make better use of cheap power. Timers for activation of clothes washers are just one example. Activation of dish washers would be workable as well.

    Take for instance freezers and refrigerators. If one can tell the freezer to make the contents go down to -10 or -20 Centigrade during the night time then come morning time it can gradually heat up without coming on. Or a separate highly insulated compartment could be designed to store extra coldness for the fridge and freezing portion. The ability to store extra coldness generated during the off-hours might work for air conditioning systems as well.

    Also, ever noticed how much power a vacuum puts out? I have a 1500W Eureka. It puts out a lot of heat and uses a lot of energy. Picture a Roomba-like automated vacuum that would start vacuuming the house in the early morning hours.

    Also, higher peak hour pricing would influence individual choices for when it becomes justifiable to install solar or fuel cell local electric power generating devices.

  13. Fifteen or 20 years ago PG&E installed a time of use meter on my residence in Pleasanton, California. We bought various gadgets to shift heavy energy use to off peak times. The program was discontinued when California froze the demand side price structure and deregulated the supply side.

  14. We work all week. We do our laundry on Sunday. I assume that would be off peak. The real problem will be to get the kids to turn off the lights when they leave in the morning:-)

  15. I have a problem with option #4. The utility has a substantial investment in the wires, transformers and poles. I have no problem with allowing customers with on-site generation to leave the system without exit fees, since the capacity which served them can be used to accommodate growth on the distribution system. However, if the customer remains connected to the grid and expects no-notice service on demand (and especially on-peak), I believe that customer should pay the utility a monthly fee which allows it to recover its investment in facilities which are still dedicated to that customer’s service.

    Having one’s cake and eating it too may sound attractive, but it is definitely not equitable.

  16. I think Ed’s right about option #4, and that it should reflect that continued connection to the grid has a real option value. We are not recommending ignoring that option value.

  17. Some further thoughts.

    The phone company (when there was one of them) use to have three rates for long distance calls. A day rate from 7:00 AM to 5:00 PM, an evening rate from 5:00 PM to 11:00 PM, and a night rate from 11:00 PM to 7:00 AM. to the best of my recollection the evening rate also was on Saturday and the night rate on Sunday until 5:00 PM. I think that when Mr. Smith and Ms. Kiesling refer to off peak rates they are thinking of the equivalent of the phone companies evening rate rather then the night rate.

    Perhaps, at least for the subject of discussion we should define Peak hours to be 9:00 AM to 9:00 PM. Sub peak to be 6:00 AM-9:00 AM & 9:00 PM to 12:00 AM. Off peak to be 12:00AM to 6:00 AM. I suspect this matches reality more then just using peak and off peak.

  18. Speaking as someone who forgets to move the wash to the dryer and as someone who has ruined a few pots and pans by forgetting something is cooking until it is burned I have to say that more sophisticated timers would be very useful. I’d love to be able to tell a stove to turn off a pot once most of the water was gone.

    I’d also like a house network connecting my computer to my washer and dryer so the appliances can tell me when to go attend to them.

    I’d also like an automated system that switched clothes from the washer to the dryer without my intervention and that could then take a pile of clothes that is backed up waiting to enter to come in and start the wash cycle while the previous load does the dry cycle.

    Smart appliances are coming. Very low power embedded processors which can hibernate and be awakened by timer clocks could wake up and start some process. Also, once the process was finished they could send a message to a house computer that is running in a very low power mode so that the house computer could tell a human at the appropriate time.

    As for houses that have their air conditioning turned off during the day: This is easy to deal with. A person could send a signal via a cellular link when about to leave the office to order the house to start cooling down or the cooling-down mode could be triggered at some previously set time.

    There is another thing that could be done with demand-based pricing: electric companies could send out a signal modulated on their power lines to broadcast rate changes. Home computers or business computers could control the level of activity of various devices based on the current electric power price.

  19. Well as somebody in the energy industry I can tell you the 2 AM line was probably coupled with a bit of hyperbole. Off peak hours are not just the dead of night. Weekends are off peak, and so is anything after 11 PM. On peak ends at 6 PM where I live. From 6 PM to 11 PM is mid peak. This information is most likely easily obtainable from any utility website. I know here in CA, it is required that the utility have an easily accessible copy of their tariffs for customers. Typically you can find pdf versions of terrifs on the websites with 3 maybe 4 clicks.

    As for seperating the lines and wires from energy, that was precisely what one utility in CA was proposing. They were going to propose a high fixed charge and very low energy charges. The argument was going to be that that was a big part of the purpose of deregulating, to get the right price signal to the consumer. Then the energy crisis hit and everything went to shit.

    The problem with residential customers is that there would be a learning period where they figured things out, such as when prices would tend to be high and what their own load profile is like so as to best adapt to the price signals. For your typical residential customer I don’t know, but it would seem to me the learning period could be quite long.

    For large industrial customers they’d probably respond to price signals pretty quickly. Some large power consumers actually have people who worry just about the electric bill. So they’d basically be getting paid to figure these things out. Failing that consultants might also find it profitable to provide them with such information.

Comments are closed.