Rationalizing California’s Residential Electricity Rates
California is finally talking seriously about changing the way utilities price electricity for residential customers. In particular, as a result of recent legislative actions, the CPUC now has some flexibility to modify the extreme increasing-block pricing (IBP) schedules that were adopted after California’s 2000-01 electricity crisis. IBP means charging more for additional kilowatt-hours as a household consumes more over the billing period. Pacific Gas & Electric’s current residential rate schedule, shown in the figure below, illustrates how IBP can charge drastically different prices for an additional kilowatt-hour (kWh) depending on how much the customer is consuming. (The rates of Southern California Edison and San Diego Gas & Electric look pretty similar.)
PG&E’s current increasing-block residential electricity rate structure
This extreme form of IBP violates one of the basic tenets of utility price setting: cost causation. Prices should reflect the cost of serving a customer. The cost of providing electricity does vary over time, an issue that I will return to below, but it does not vary with how much a household uses over the billing period. That kilowatt-hour (kWh) for which PG&E charges one customer $0.32 costs the utility the same to provide as a kWh for which PG&E charges another (or the same) customer $0.15.
The implementation of IBP in California is both unfair and inefficient, a holdover from policy made during the darkest days of the electricity crisis without careful analysis of the impact. While California’s IBP schedules are more extreme than in other states, the lessons from the Left Coast have much to teach in other locations as well.
To remind the young or forgetful (or the less-obsessed with electricity tariff history), prior to the electricity crisis, the regulated electric utilities in California had two-tier IBP rates with the second tier 15%-20% higher than the first tier. The majority of customers consumed more than their first-tier baseline quantity for their region, paying the first-tier price up to the baseline and the second-tier price for any additional electricity beyond that. But more than a third of customers only experienced first-tier rates.
After the crisis, utilities needed to raise revenues, but politicians said they wanted to protect the poor, so regulators instituted a 5-tier system, with rates on the first two tiers – designed to go up to 130% of baseline quantity, which is about the median household’s consumption – frozen at pre-crisis levels. As a result, any revenue increase had to come on tiers 3, 4, and 5, which constituted only about one-third of the residential electricity sold. (Tiers 4 and 5 were merged into one tier a couple of years ago.) To get the needed revenue, the price on those high-tier kWhs had to go up a lot.
Southern California Edison’s tiered rates from 1999 to 2009 (Source Ito (2014))
That is how we got to the IBP schedules we have today where the highest-tier rates are more than twice as high as the low-tier rates (though the high tiers change frequently, so the exact ratio is a moving target). This ratio is far out of line with other retail rate structures. The majority of utilities in the U.S. don’t have IBP; they sell all electricity to residential customers at the same cost per kWh. In the one-third or so that do have IBP, the rate structure typically looks more like pre-crisis California, with two tiers and only a small difference between them.
Lots of claims get lobbed into the political heat of the discussion about IBP, generally with little empirical basis, so it is worth reviewing the best evidence we have. Increasing-block pricing proponents make three cases for IBP:
First, some proponents argue that high-usage customers are more expensive to serve on a per-kWh basis because they on average consume a higher share of their electricity at peak times. It is ironic that some of the same organizations that make this argument also are the adamant foes of time-varying pricing that would reflect time-varying costs directly. In any case, my own research (published in Review of Industrial Organization in 2013) shows that there is a tiny bit of truth to this claim, but the operative word is tiny. That study shows that for PG&E and Southern California Edison, the different time-patterns of consumption between large and small residential consumers could justify at most about a few percent price difference. That would be less than a one cent per kWh average price differential, far smaller than would result even if we returned to the pre-crisis IBP structure, and vanishingly small compared to what we have today.
My study shows that the true incremental cost per kWh of serving large and small residential customers is virtually the same. And that cost is much lower than the upper-tier prices under the IBP used by the California regulated utilities. Even if you added the carbon cost of the emissions and priced those emissions at $37/ton (the latest estimate of the social cost of greenhouse gas emissions from the Obama administration), that would raise the appropriate price by less than two cents per kWh.
The second argument for IBP is the one that drove the political debate in 2001, that IBP is a way to raise rates while protecting the poor. Implicit in this claim is that poor people consume less electricity than rich people do, which makes intuitive sense, but at the time the legislation was rushed through during the crisis there was just a sense, not a measure of the real difference. A paper that I published in 2012 (in American Economic Journal: Economic Policy) shows that using the steep post-crisis IBP does help low-income customers relative to a flat rate, but by much less than you might think. The reason is that the CARE program that is explicitly designed to give lower rates to low-income consumers – those up to 200% of poverty or $47,700/year for a family of 4 – has already taken most of the low-income customers out of the standard IBP tariff and put them on a separate low-income tariff.
Overall, I estimate that moving to a flat-rate tariff would raise the average bill of a customer making about $50,000/year or less (in today’s dollars) by about $5/month. That’s not nothing, but it’s also not a huge effect to get in exchange for an increasing-block rate structure that greatly distorts prices away from real costs – which are essentially the same for the low-tier and high-tier kWh. Furthermore, moving from IBP to a flat rate would likely incent more truly needy households to sign up for CARE. If the goal is to help the poor, a program targeted directly at the poor makes more sense than one targeted at consumption level, which has a weak correlation with how wealthy a household is.
Using IBP to help low-income households has another problem. It arbitrarily harms many low-income households that are heavy users for completely understandable reasons. Household baselines are based on climate regions (PG&E has 10 regions, SCE has 6) and…well…nothing else. They don’t adjust for how many people live in the house, how old those people are, how much time they spend at home (rather than using electricity somewhere else), or any other reason that we would expect even an energy-conscious household to consume more. Instead, IBP rewards people who live alone and don’t spend much time at home. Does that make sense?
The third case made for IBP is that it leads to conservation, because it raises the price of marginal consumption. In theory, this could be right, but in practice it doesn’t work out that way. The idea is that it raises price for the heavier uses and lowers price for the lighter users, compared to a single flat rate, but the response of heavy users is proportionately larger.
But path-breaking research that our former grad student Koichiro Ito published this year in the economics profession’s top journal (American Economic Review) shows that most customers don’t pay attention to the marginal price they face, but instead respond to the average price or total bill. This isn’t surprising given that the vast majority of customers don’t even know we have tiered pricing and even for the electricity-obsessed it’s very hard to know what marginal price you will face at the end of the month. Ito demonstrates that when customers respond to average price, IBP results in virtually no change in total consumption.[1]
Residential Solar PV has been a big beneficiary of increasing-block pricing (Source:http://256.com/solar/images/)
But what has become apparent in the last few years is that one small set of customers is very aware of IBP and the high marginal prices for heavy consumers: solar PV households, or actually the solar PV vendors who explain IBP to them. While evidence that IBP incents energy efficiency is absent, the evidence is clear that IBP is a key driver behind the distributed solar PV movement in California.
In fact, solar PV installers “work with” customers to optimally design systems so they just shave off the high-tier usage without touching the low-tier usage where the price is way too low for PV to save customers money. With federal subsidies that pay nearly half the cost of PV — and net metering that pays the residential customer for electricity supplied to the grid at the retail rate rather than the wholesale rate that other renewable generation sources receive – solar PV can beat the high-tier prices of IBP. That’s why solar PV installers are the most vocal opponents of rational rate reform that would call a kilowatt-hour a kilowatt-hour regardless of how many other kilowatt-hours you consume during the month.
So, increasing-block pricing isn’t cost based. It is a very indirect way to lower the average bills of low-income customers, while arbitrarily harming many of them and benefitting many wealthy people who live alone or don’t spend much time at home (or own a second home). And it’s only effect on conservation is to create incentives to install solar that just skims off the highest tiers of the IBP structure, an activity that is pursued overwhelmingly by the richest households (and which raises the bills of all other ratepayers). The first step to rationalizing California’s electricity rates is to greatly reduce or eliminate increasing-block pricing.
Then we need to talk about time-varying pricing. We should be transitioning to opt-in time-varying pricing immediately and to opt-out in the near future. I’ve explained an efficient and equitable way to do this in a recent paper, but that will have to wait for another blog.
[1] Paul Chernick recently filed testimony on behalf of NRDC that mischaracterizes Ito’s finding, saying Ito claims consumers don’t respond to IBP. That’s not at all what Ito finds, as I explain here. A study by Energy & Environmental Economics argues that in British Columbia IBP does lower total consumption. But the study looks only at whether high-consuming households consume less under IBP – which they do, not surprisingly—and ignores that low-consuming households consume more, because their price is lower than under a single flat rate. What matters, of course, in the change in total consumption, on which the E3 study is silent.
I’m still tweeting interesting energy news articles most days @BorensteinS
Categories
Severin Borenstein View All
Severin Borenstein is Professor of the Graduate School in the Economic Analysis and Policy Group at the Haas School of Business and Faculty Director of the Energy Institute at Haas. He received his A.B. from U.C. Berkeley and Ph.D. in Economics from M.I.T. His research focuses on the economics of renewable energy, economic policies for reducing greenhouse gases, and alternative models of retail electricity pricing. Borenstein is also a research associate of the National Bureau of Economic Research in Cambridge, MA. He served on the Board of Governors of the California Power Exchange from 1997 to 2003. During 1999-2000, he was a member of the California Attorney General's Gasoline Price Task Force. In 2012-13, he served on the Emissions Market Assessment Committee, which advised the California Air Resources Board on the operation of California’s Cap and Trade market for greenhouse gases. In 2014, he was appointed to the California Energy Commission’s Petroleum Market Advisory Committee, which he chaired from 2015 until the Committee was dissolved in 2017. From 2015-2020, he served on the Advisory Council of the Bay Area Air Quality Management District. Since 2019, he has been a member of the Governing Board of the California Independent System Operator.
Energy Institute Blog 
Here’s the quote from the abstract in Ito’s AER paper: “Using household-level panel data from administrative records, I find strong evidence that consumers respond to average price rather than marginal or expected marginal price.” Seems quite clear–marginal price conveyed through the IBP is no more effective than average pricing at inducing total conservation.
Ito and Borenstein are for TOU and RTP rates if you’ve read other posts on this blog. They are arguing that IBP is no better than flat rates, but TOU/RTP is preferable to flat. TOU/RTP would given the needed incentive to reduce peak demand. (And see my other post about using a daily demand charge as an alternative.)
As for the equity effects of the IBP rate, Borenstein has written a paper showing a weak relationship between usage and income (he references it), but it still creates a large amount of policy “leakage”–wealthy individuals gaining from a program aimed at the poor. The most obvious fix is to simply target the CARE program better. There are MIUCH more effective means of achieving those equity goals. (And load up conservation goals onto a very small segment of the population is pretty stupid and leaves much of the most cost-effective conservation undone. In fact, IBP encourages lower consumers to use MORE power.)
I do know which pricing method is absolutely wrong–one that charges increasing prices as load accumulates during a month. That form is totally contrary to the principle that electricity is a non-storable good. With smart meters its a truly ridiculous approach. Electricity pricing isn’t as complex as you make it out to be based on my experience.
“Weeds” reminds me of the old adage “Don’t sweat the details. Only problem – it’s all details”.
I’ll recap my points: The Ito paper quoted by the OP Borentsein and mcubedcon does not actually support the conclusions they claim or infer. It may be interesting and clever research, but claiming data supports something it doesn’t is a favorite tactic of those with a political agenda. Ito is very specific about what the evidence (data) show and I am taking his claim at face value. He then conjectures some reasons why marginal pricing has no elasticity of demand, which he is entitled to do. But even he does not claim the evidence proves his conjectures – that is for further research.
Despite all the hand waving above nobody has retorted by quoting specific lines in Ito’s work or referenced research as I have.
Now back to the general debate about IBP, which is essentially a political debate. We need to step back and remember the general idea and objective of IBP as implemented in California after the electricity crisis a decade ago. One goal was to promote some conservation, but target it more at the highest users which approximately translated into the richest users since they tend to have bigger houses that use the most. And yes, as pointed out in the OP, unwittingly a few poorer users may have been caught in the net. Nobody said this would deliver the most total conservation despite inference of OP. You can’t fault a program for not meeting a goal it never had. Obviously the way to promote the most total conservation is to double everybody’s bill ((flat rate) and you would get conservation. But that was not politically feasible, since it was felt that richer people could better afford the investment in efficiency and bear the costs of higher rates. So IBP delivers a much higher average cost to the outliers on the distribution curve and surprisingly seems to have resulted in a little conservation in those targeted.
Ito and Borenstein are clearly for flat-rate and they both say that, and that is their opinion. But the research quoted does not address the original problem which is how to provide revenue to build peak capacity to avoid brown outs. Any rate change that increases total revenue would help but they have not proven that a small flat rate increase would lead to substantial conservation statewide in California. A large flat rate increase is politically untenable, and even a small one would be difficult.
I used to be against IBP and for 5-minute increment TOU pricing with technology that allows low information and transaction costs (classic efficient markets) for both supply and demand sides. But as I learned more and more about the complexity of the electrical system and pricing I no longer think it’s such a panacea. Electrons are not like apples so our everyday supermarket experience with commodity pricing does not apply. That leaves me agnostic about whether IBP, flat rate or 5 minute TOU is best. Maybe something simple like policy that promotes the retirement of the oldest most inefficient gas burners replaced with new and shiny ones may yield a nice mix of pollution reduction, cost and reliability. Not that sexy mathematically, but could be just fine.
I looked through your previous comment and Ito’s paper and I can’t see what you’re claiming Ito’s paper doesn’t prove. Section 3 of his AER paper goes through 3 empirical tests of whether marginal prices have price effects beyond the average price. It doesn’t have the “silver bullet” sentence you may be looking for, but that’s often the case in a journal article. So maybe if you’re more specific about what it’s missing I can address that.
Perhaps we all need to come up out of the weeds a bit and look at a bigger piece of the picture. There seems to be a presumption that electricity prices will be set by a regulator rather than as a result of interactions between buyers and sellers, and that pricing will consider or be influenced by a sizable number of non-market factors, like equity, cost causation, renewable energy policy, and carbon abatement (which is only weakly influenced thus far by renewable energy policy), among others. If I’m correct, then I think it’s fair to say politicians have created something akin to a mathematical programming problem in which there are too many constraints to allow an efficient solution, let alone an optimal one. Before we launch into a detailed rate design discussion, I think we have to back up and talk about scaling back the extent to which energy policy is used to achieve other objectives. If we’re going to focus on equity, then efficiency is a lost cause. if we’re going to focus on efficiency, they we need to be prepared to sacrifice equity to a large extent (and help low income consumers in another way). It’s my observation after more than 40 years in this business that by trying to do too much, we’ve actually accomplished very little.
Two points. I do agree that we try to do too much with energy policy, such as “cobenefits” from AB 32 measures that undermine the true purpose of the program. However, that said:
– While it would be great if we could conduct policy analysis as you suggest, that simply isn’t going to happen and what we say here won’t change that fact. We have to play with the cards we’re dealt. That means we’re focused on residential rate design.
– There are ways of addressing equity in a trade off with efficiency. But we can talk about how to address equity while avoiding distorting efficiency, e.g., giving a monthly lump sum payment instead of a rate discount, which is what SMUD does.
Again I think people are missing the subtle but important point of MC. The Wellington NZ idea is logical because it tries to approximate true cost as FC + MC. But the term “wholesale price” opens a can of worms in the Golden state. My understanding of how wholesale power is bid on CAISO is that it bids hourly in the day ahead market and 15-minutes in the real time market and is dispatched in 5-minute increments. http://www.caiso.com/market/Pages/MarketProcesses.aspx
But the subtle point is that they take the highest bid as the market clearing price and apply it to _all_ suppliers for that period. Thus wholesale price quickly deviates from actual cost, except maybe for the last most inefficient generator that was dispatched. It’s counterintuitive but makes sense when you really dig into it.
As a semiconductor technologist I thought it would be cool to implement a solution where every appliance in the home whether at the wall or in the fuse box meters 5-minute data. Technology is already there. Combine that with 5-minute pricing (firmware upgrade to smart meters that have already been widely installed) and voila. But as I learned more about all the issues I realized sadly that it is very hard to figure out true cost at the peak, other than maybe the cost of actual fuel burned which will yield surprisingly little peak MC variation when averaged across all consumers I have been told by those with access to the inside data.
Another way to think about this is to consider the last 1% of capacity. As the system dispatches the last 1% which by definition are the oldest most inefficient generators, since electrons are fungible consumers would hardly notice the inefficiency if you passed on the direct fuel cost because it is blended with the other 99%. But when you hit 101% you notice the step function as brown outs start. The purpose of peak TOU or 5-minute pricing however it is proposed is to scare people away from the 100% so it needs to be punitive, not actual. Which brings us full circle … you guessed it …. to IBP. I bet if the 5-minute thing is ever implemented they would try something like 5-minute IBP.
Markets work by pricing at the marginal cost of the highest cost producer. Prices reflect VALUE, not the average of supplier COST. From an economics standpoint, CAISO is correctly pricing, at least conceptually. (There are other issues I’ll discuss.) The price increases i a nonlinear fashion as the cost approaches peak capacity. There’s no need for an additional “penalty” (along with the fact that the reliability threshold is always changing hour to hour.) Thus IBP is basically unneeded and even extremely unwieldly for the purpose you describe.
As for CAISO’s wholesale price, the problem is that the value of capacity is not included because the largest players in the market recover their capacity payments through separate payments outside of the DA and RT markets. Utility owned generation is ratebased, and much of the merchant fleet is under bilateral agreements. The DA and RT markets are actually surplus dump energy markets and don’t reflect true value.
“consumers would hardly notice the inefficiency if you passed on the direct fuel cost because it is blended with the other 99%”
A last 1% unit’s possibly very high fuel cost is not blended with the other 99% if it’s a marginal unit that sets the clearing price..IF it sets the clearing price; units can also be dispatched outside of the market, where the cost is spread evenly (I think) to everyone as “uplift”
If you want to scare people away from using up that last 1% of capacity, look at Texas; next year there will be a $9000/MWH offer cap. The CAISO cap is $1000/MWH.
The economic case for the block pricing of water (especially groundwater) is to face consumers with the full marginal cost (FMC) of water (on marginal units) without making excess profits. This is because the FMC is increasing in quantity (also in distance/elevation). As Vernon Smith has long observed, generation and transmission of electricity are also subject to rising MC, so the same thing should apply in electricity. Adding time-of-day and space is important, but doesn’t change the fundamental motivation for block pricing.
If consumers are not equating MB w/ price, the remedy may be presenting the information in a more dramatic fashion.
Regarding solar, the devil is in the subsidy, not block pricing.
The difference between water and electricity is the ability to store. Water used at different times in the month, or even the season (e.g., summer), have similar costs, with the increasing marginal costs you identify. In the electricity market, there are 8,760 hourly markets that open and close each year. There are important intertemporal elements in the market, but each hourly market essentially starts anew. So the increasing marginal costs are largely contained within a single hour. So if there is block pricing of electricity, it has be with a single hour. (You might be able to stretch to a single day if you include unit commitment and hydro storage costs.) This is the fundamental difference between water and electricity markets.
The only reason for starting w/ water is to understand the economic basis for increasing block pricing in the first place. A case can be made that if you cannot price electricity according to actual fluctuations in MC, an increasing schedule (with only two tiers) is preferable to a single rate, but of course still far from first best given the tremendous variation of MC over time. One of the retailers in Wellington, NZ charges customers the time-of-day wholesale price plus a distribution charge. That’s still not first best but getting pretty close.
The problem with time-varying pricing for electricity idea is to determine the cost at peak timeslots. Contrary to popular belief the true cost has little correlation with the marginal fuel cost for that peak time slot. Most of the cost is the capital infrastructure needed to purchase and maintain peak generation and transmission capacity that is “off” for most of the time. In a complex distributed system it is often difficult to separate out those costs in a meaningful way for many reasons.
Even if we wave a wand and assume 5-minute smart metered and smart appliances are installed statewide, consequences are hard to predict. The closest practical analogy to peak pricing I can think of is Donald Shoup’s experiment with SFpark. While it has had some positive results, an unexpected outcome was the lowering of total system revenue, something that neither his model nor the detractors predicted. The reasons are still under investigation.
I would also like to challenge the assumption that CA residential consumers don’t take IBP into account in their technology decisions. While that is a reasonable conjecture given how confusing IBP in CA is, I would like to see a current comprehensive statewide survey of residential bill-payers and decision-makers to prove that. The referenced research seems to be nation-wide, but I think CA is the largest and most unique case given the extremes of the tiers. I for one made a conscious decision not to install air-conditioning because it would bump my low rate into a much higher tier. Instead I use a whole-house fan which uses a fraction of the energy and is about 90% effective for where I live. I recognize some consumers may decide they need to be cool every minute of the day no matter the cost, or may be in a warmer climate zone, but the conclusion in those cases may be that flat-rate or IBP would be a don’t care for their investment decisions, not that flat-rate would yield a more favorable outcome however that is defined.
Yes IBP has a tendency to offend my non-market pricing sensibilities, but my point is that it is not so simple to design a rational alternative in practice for something as complex as the electricity grid.
Having worked on rate design at the CPUC for a couple of decades, it’s really not that difficult to identify the higher and lower cost time periods. The current problem is that time periods haven’t changed in a couple of decades, and they probably should given the change in conditions. However an important issue to address is how to protect consumer investments that were made in response to the current TOU periods. I’ve written about that issue here:
http://mcubedecon.wordpress.com/2014/07/01/rethinking-the-rates-that-utilities-offer-to-customers/
In testimony submitted to the CPUC last year, we showed that TOU rates would lead to lower revenues, but also that they would lead to even lower costs–it needs to be a two-sided equation. Any analysis of time varying pricing that takes the current infrastructure as static and ignores the effects on future investments is missing the whole point of time varying pricing. I’m disappointed that Shoup’s analysis didn’t consider this aspect if that’s the case.
Ito’s paper referenced here looked directly at SCE’s and SDG&E’s block rates over about a 7 year period. It was California specific and demonstrated empirically and convincingly that residential customers are largely unaware of which tier they are on. I strongly suggest reading it.
I have read Ito’s paper. He only looks at about 40K households in a few cities around Laguna beach / Mission Viejo. I live in CA I know the area I looked at the map. There are 12M households in the state. That would be a tiny fraction. On pg 537 he states “Many surveys find that few people understand the marginal rate of nonlinear price” and the references 2 and 3 seem to be nationwide not CA-specific statewide (or very old).
He states in several places “I find strong evidence that consumers respond to average price rather than marginal or expected marginal price”. He uses the term “The evidence strongly suggests” they respond to average price. He then proceeds to conjecture of “likely” reasons why, and understanding (information cost) is one of three possible reasons – but he does not claim any reason is proven by the evidence (data). So your conclusion above that “customers are largely unaware” is not an evidence-based conclusion of this paper. If there is other research that proves that CA statewide I am interested to know which.
To me there is no surprise that there is no elasticity of demand to marginal pricing. That outcome I have predicted before any knowledge of this paper. I am actually skeptical of his result that there is elasticity of demand even to average pricing. I could see that for large changes in average price but not for moderate changes. Yet he does seem to show that result. The strange part is you would expect the demand change to lag the bill a few months but he did a lag analysis and showed no difference. That is surprising and that is why I’m skeptical.
What I could not tease out from his paper was how he separated marginal from average. What IBP does is deliver an exponentially high average price (ie high monthly bill) to users that deviate from the mean. By definition the average encompasses a marginal component. I know he has an encompassing test and bunching analysis but it was not clear to me exactly how he separates the marginal component. It also wasn’t clear to me the exponent of the average price elasticity curve – IE if you just had a small flat rate increase rather than IBP would there be a noticeable elasticity. So I think he proved no response to marginal but that is not the same thing as proving an elastic response to average. It’s a subtle point the difference to betwen response to average vs response to IBP since IBP delivers a quasi-exponentially rising average.
The reason people don’t respond to marginal is I think obvious, and he suggests it as his second reason: You don’t have enough information during the month to fine-tune your consumption to bunch at the tiers. All you know is you received a shocking price signal the previous month and you try to “use less” by cutting where you can, raising your thermostat etc. Your response it is not possible to judge how exactly how far to cut and when to stop cutting. Technology could change that but I still think we will get unexpected results even if the magic wand is waved and we have 5 minute per-appliance metering and pricing all neatly controlled by your smartphone.
You need to understand how economic and statistical analysis is done. You do not need to analyze all 12M households to gain insights into how they will respond. The 40,000 household is an extremely large data set for economic analysis, which is one reason why it is so powerful. There is no evidence that consumers in CA are any better informed about electricity price than elsewhere. The burden would be on opponents of the paper to produce updated data.
You’re getting into semantics. Read other academic papers. NONE of them ever claim that something is absolutely proven. They always point to weight of the evidence. IBP support is equally weak if you’re trying to make that point.
You might be skeptical of the results, but I suspect that you are not very familiar with the large literature on energy demand. Start with Carol Dahl’s ongoing survey of energy demand elasticities.
You need to understand the econometrics to see how he derived his results. They are quite powerful in that context.
Nice paper. But since most customers don’t realize IBP exists, why not use another rate form that most customers don’t realize exists, a demand charge. A demand charge will collect revenue from those customers who impose costs on the system in terms of the utility having to own the wires that go to their homes, as I stated in “Curing the Death Spiral,” with Lori Cifuentes (Tampa Electric Company), Public Utilities Fortnightly, 2014 August.
https://www.fortnightly.com/fortnightly/2014/08/curing-death-spiral?authkey=54d8da5efd3f76661023d122f3e538b4b3db8c8d5bf97a65bc58a3dd55bb8672
There are various forms of demand charges available. I remember a funky one I saw used in Burbank almost 40 years ago.
I think residential demand charges could quite constructive, but with a very important caveat–they must be DAILY demand charges, not monthly. Monthly charges penalize a misstep in a 30-day block and obliterate any conservation incentive once the monthly peak is hit. A daily charge is much more forgiving and resets the incentive each day. With advancing metering technology is is easily feasible now.
Considering that the customer wants the utility investment to be there for a once a year event, or a once in ten year event, and that the utility makes such investments for an expected life that is even longer, a daily demand charge is too short.
Let’s begin with the premise that a “month” is an arbitrary time period invented by people to observe religious holidays based on cycles of the moon. There’s nothing magical about a monthly billing period other than tradition. You are correct that the annual cycle is what drives the peak demand. However, the annular peak does not occur on the same day each year–it occurs somewhat arbitrarily within a 4-5 month period between June and October. That annual peak is actually the result of the high on a single day when the aggregated demands of all customers daily peaks reach a system high. The annual peak is NOT the sum of the monthly peak demands of all customers–in fact those peak demands have to be statistically adjusted for relative coincidence compared to the actual single day peak demand. Moving to a peak demand charge that includes the probability of that type of day being the annual system peak is a much more efficient pricing signal, plus it gives a new conservation signal each day. The monthly demand charge fails miserably at this, and also creates all sorts of distorted actions at the start and end of the billing cycle. An annual demand charge is little different from a fixed charge and shouldn’t even be on the table.
I don’t see how James Rousasset (Mcubedecon) justifies equating an annual demand charge with a fixed charge, which I presume means a customer charge. Demand charges change with the size of the customers need/desire for service. A fixed charge wouldn’t and would just drive up the cost of the small customer.
The most important argument against an annual demand charge–customers would have to keep track of what had been their peak demand throughout the year. And once a customer has hit the peak demand for the year, e.g., had on their AC, refrigerator, microwave et al simultaneously, they would have NO incentive to avoid hitting that peak demand again over and over throughout the year. (This already happens with monthly demand charges on commercial customers.) The fact is that the electricity system can be more stressed under lower peak demands, e.g., during a drought when hydro generation is less available in August and September (which is what happened in 2000.) Because of the single event nature of the demand charge and the likelihood that it will be quite similar year to year, it starts to look a lot like a fixed customer charge, being largely invariant. (You’ll have to provide empirical data that shows that residential peak demand varies substantially from year to year for individual customers.)
An added problem, particularly in the PNW, is choosing when to start the new year. January 1 is in the middle of the peak load season (which is actually energy driven rather than peak hour driven), but the PNW may be pricing on a WECC-wide basis which is summer peaking. Daily demand charges avoid all of these issues.
The argument against a demand charge seems to be generation driven with the reference to hydro resources causing stress despite California being largely a restructured utility environment with the utilities who would implement the demand charge largely not responsible for generation. The longer time period would collect more money from those customers who are using the system on a standby basis to backup their solar. This changes the economics of installing solar from being against the bloated energy charge to being against an energy charge with some of the fixed costs pulled out and collected as a demand charge, which the solar customers will have a harder time bypassing.
An interesting aspect is that solar customers might actually benefit from a well-designed demand charge, and depending on what is defined as the “peak” period. Is that peak period before OR after the introduction of a large amount of customer-owned solar power? This is the same question that California faces when calculating avoided costs for QF: Is the incremental price equal to the difference between the QF In/QF Out costs, or to the marginal cost of QF In? For TOU pricing, is it metered load or plug load that matters as they diverge. The policy answer is that it depends on how the CPUC wants customers to act in the future.
Electricity, unlike water, must be consumed ‘instantaneously’, so the ‘price’ must have a tighter relation to cost. Thus time-based-consumption pricing indeed makes sense.
IF I am to be charged a time-of-use-based rate, I must also know the rate I am being charged at any instant. Now that a larger % of utility customers have ‘smart’ meters it should be quite easy for the consumer to know how much kwh-rate they are consuming at any instant, and thus know what it is costing them. There must already be devices and apps that can convey this info to the ‘ratepayer/ customer’. Automation can then take care of optimizing the consumption to the customer’s ‘liking’.
Why the reference to water: same problem as any utility. I just moved from a municipality where the charge for water usage from the lowest to the highest is about $2.75 to $3.45 per ‘unit’ to an area where the difference is $3.25 at the lowest tier to $10.75 at the highest. THAT kind of difference gets one’s attention; every long shower costs $10.75 [a unit is 775 gallons], and even worse watering your lawn in the East Bay can set you back about $15/day > NOW you can decide if the green stuff is worth $450/mo.