Risks of Going Solar

On February 22, 2016, Catherine Wolfram posted the blog Risks of Going Solar on the Energy Institute at Haas blog, part of the University of California Berkeley.  I posted the following, which I am adding to my blog.

Of the various regulatory Risks of Going Solar, Catherine Wolfram identifies two biggies, reducing the size of the net metering interval and shifting the rate design to include a smaller energy charge and a greater fixed charge.  But the risk of these two can be much larger than Dr. Wolfram suggests.  Reducing the size of the net metering interval exposes rooftop solar customers to the possibility of negative prices, while cost re-classification could result in (greater) demand charges instead of greater monthly customer charges.

In “Renewable Electric Power—Too Much of a Good Thing: Looking At ERCOT,” Dialogue, United States Association for Energy Economics, 2009 August,[1] I point out that a surplus of wind in West Texas forced the wholesale price for electricity below zero for about 25% of the pricing periods during that April, at least in West Texas.

Transmission constraints generally kept these negative prices from spreading to the rest of Texas.  Negative prices did spread to other parts of the state for just less than 1% of the rating periods.  As Dr. Wolfram well pointed out, these pricing periods are sometimes as short as 15 minutes (as they were in West Texas at the time), though are often one hour.

Many ISO do not seem to allow prices to go negative.  In West Texas, the combination of transmission constraints and the various credits[2] given to wind led to negative prices.  I believe that similar combinations elsewhere will force ISOs to allow negative prices in their dispatch programs.

I have long seen the need for utilities outside the footprint of an ISO to implement real time “value of solar” prices that are similarly negative.  Hawaii seems to be ripe for such negative solar prices.  Utilities outside the footprint of an ISO can implement “value of solar” prices using a Walrasian auction, as is discussed in many of my articles.

I actually disagree with the concept of a separate price for “value of solar.”  If we are to use prices to influence generation, there shouldn’t be a separate price for solar versus other spot generation imbalances.  A different price for unscheduled versus scheduled generation, yes, but not a separate price for just solar.

There will often be many prices during any pricing interval.  For instance, a single 15 minute period may be part of a 24×7 contracted delivery of power with one price and part of a 16×5 contracted deliveries with another price.  A third price might be applicable to variances.  Variances would include both solar that is dumped into the system and hiccups in the 24×7 or 16×5 deliveries, whether the hiccup is positive or negative.

Utility rate making often includes the concept of cost classification, where costs are identified as energy related, customer related, and demand related.  In the context of Risks of Going Solar, customer related and demand related are combined into the concept of a fixed charge.

The discussed increase in the monthly charge is only one way to reduce the energy charge.  The other way, and I believe a better way, to decrease the energy charge is to increase the demand charge, or to implement a demand charge when there is not a demand charge in place.

Customer charges impose greater burdens on small, often lower income, residential customers, while demand charges tend to protect these smaller customers, as is discussed in

  • “Curing the Death Spiral,” with Lori Cifuentes (Tampa Electric Company), Public Utilities Fortnightly, 2014 August;[3]
  • “Demand a Better Utility Charge During Era of Renewables: Getting Renewable Incentives Correct With Residential Demand Charges,” Dialogue, United States Association for Energy Economics, 2015 January;[4] and,
  • “Fairly Pricing Net Intervals While Keeping The Utility Financially Healthy,” 48th Annual Frontiers of Power Conference, cosponsored by The Engineering Energy Laboratory and The School of Electrical and Computer Engineering, Oklahoma State University, Stillwater, Oklahoma, 2015 October 26-27.[5]

Thus, as we see a continued growth in solar, I see a growing need for finer pricing intervals and a growing need for demand charges.  Fortunately, the huge growth in interval meters allow these better rate designs.  We just need to political will to implement something other than a monthly charge for energy.

[1] http://livelyutility.com/documents/USAEE-ERCOT%20Aug%2009.pdf

[2] such as production tax credits and renewable energy credits

[3] https://www.fortnightly.com/fortnightly/2014/08/curing-death-spiral?authkey=54d8da5efd3f76661023d122f3e538b4b3db8c8d5bf97a65bc58a3dd55bb8672

[4] http://dialog.usaee.org/index.php/volume-23-number-1-2015/271-lively

[5] A copy is available on my website, www.LivelyUtility.com.

Billing Tweaks Don’t Make Net Metering Good Policy

Severin Borenstein published a blog on 2016 January 4 with the title of “Billing Tweaks Don’t Make Net Metering Good Policy.”  The entry reminded me of a presentation I had made in October at Oklahoma State, so I added the following comment to Severin’s Haas Blog.

Net Metering can be Good Policy for the recovery of some costs incurred by a utility in serving its customers, such as the cost billed to it by its ISO supplier.  But for the rest of the costs incurred by the utility, such as the cost of wires and meters, a demand charge and a monthly customer charge are more appropriate.

For the cost billed by the ISO supplier, the metering periods need to be aligned, an issue that is current before FERC in Settlement Intervals and Shortage Pricing in Markets Operated by Regional Transmission Organizations and Independent System Operators, FERC Docket No. RM15-24-000.  If the ISO is billing the utility based on fifteen minute intervals, it might be good policy for the utility to bill the retail consumer for those ISO costs on fifteen minute intervals using net metered amounts during those fifteen minute intervals.

This net metering paradigm seems to be only appropriate for the charges coming to the utility from the ISO, as I discussed in “Fairly Pricing Net Intervals While Keeping The Utility Financially Healthy,” 48th Annual Frontiers of Power Conference, cosponsored by The Engineering Energy Laboratory and The School of Electrical and Computer Engineering, Oklahoma State University, Stillwater, Oklahoma, 2015 October 26-27.  A shorter version of this paper was published in Dialogue, United States Association for Energy Economics, 2015 September 1.  The full paper is on my web site  in the library under Conference Papers.

The majority of the cost incurred by an increasing number of utilities are incurred for wires.  A much better way to recover the cost of wires is a demand charge.  When the customer wants to have access to a specific amount of power, the customer can contract for wires access in that amount, which would be billed monthly based on contract demand.  Customers with poorer information about their power requirements can rely on a demand charge based on the interval with the highest net metered amount, generally fifteen minutes or an hour, though I have seen the interval being an entire summer month.  Customers who exceed their contract demand would pay for the excess demand through a multiple of the demand charge.

There are a few appropriate demand metrics, such as the customer maximum demand or more exotic demands such as the contribution to the distribution system peak or the peak on a subsection of the distribution system, all as discussed in the above paper.  We are still several years away from real time pricing of the distribution system, as I discussed in “Dynamic Pricing: Using Smart Meters to Solve Electric Vehicles Related Distribution Overloads,” Metering International, Issue 3, 2010.

The Goldilocks Dilemma

An old posting about why intermittency is not a big deal came to my attention today.  I re-read some of what had been said, especially when I had just sent out a paper on the topic yesterday.

I believe that the value of electric “energy” is often overstated.  The author of the old posting, Chris Varrone, inadvertently acknowledges this when he wrote

However, the energy in wind is worth 100% of the energy in nuclear (or anything else) in the spot market; wind energy in the day-ahead market may be worth a little less, but this can be “firmed” using energy trading desks or by using other assets in the operator’s fleet.

If the day to day differential can be handled by firming with other assets, then the value of the electricity is not just energy.  It is not worth debating what to call this other value, but a substantial part of the value in the spot market is something other than energy.

As to The Goldilocks Dilemma, the paper I sent out yesterday, I began by asking

Is the price paid to dispatchable generation too high, too low, or just right for intermittent generation?

I then answer

Though intermittent generators often argue that they should receive the same price as dispatchable generation and some utilities argue that they should pay less to intermittent generators, sometimes intermittent generators should face a higher price than dispatchable generators, such as when intermittent generation is part of the market during instances of extreme shortage.

The entire paper is available on my web site, the companion to this blog site.  Look for the hot link to the library near the bottom of the first page.  A hot link for the article is near the bottom of library index in the section called drafts.

Electric Demand Charges: A Lesson from the Telephone Industry

The only ad with prices that I remember from 50 years ago was AT&T’s offering of a three minute coast to coast telephone call for $1.00.  With the inflation that we have seen over the last 50 years, one would expect that a coast to coast call would now be at least $10.00 for three minutes.  Instead, most telephone bills show a monthly service fee and no itemization for individual calls.  Automation has allowed the telephone companies to do away with most telephone operators, which was a significant portion of the variable cost of making long distance telephone calls.  The principal cost is now the investment in the wires, which doesn’t change with the number of calls that are carried.  So, most carriers now charge a monthly fee and little or no charge per call.  Perhaps it is time for the electric industry to go that way?

 

The restructuring of the electric industry has generally separated the distribution wires function from the generation[1] and transmission[2] function for most customers of investor owned electric utilities.  This restructuring puts such electricity customers into the same position as their counterpart customers of municipally and cooperatively owned utilities.  Municipally and cooperative owned utilities have generally been distribution only utilities, buying generation and transmission services from others, instead of being vertically integrated like most investor owned electric utilities.

 

The restructuring of the electric industry has resulted in most customers being served by a distribution company which has very little variable cost, much like the telephone companies.   A significant distinction is that telephone lines handle one call at a time.  The telephone line is either in use or is not in use.  In contrast, electric utilities provide a continuously variable service.  The customer may be taking 10 watts (a small light bulb) or 10 kilowatts (running the A/C, water heater, and stove at the same time), or any amount in between.  The telephone company has the wires to serve the customer’s demand, whenever that call occurs[3].  The electric distribution company similarly has the wires to serve the customer’s demand, whenever that demand occurs.  While the telephone company will have customers on a binary basis (they are either a customer or are not a customer), the electric distribution customer serves its customers on a continuous basis (they might be very small customers who never use more than 10 watts or a very large customer that might use up to 100 MW.)

 

The binary basis of telephony customers allows the telephone companies to charge their customers a specific amount on a monthly.  The continuous nature of the size of electric services suggests that electric distribution companies charge their customers a price based on the size of the electric service used by the customer.  For commercial and industrial customers, electric utilities have long included in their tariffs a demand charge that depends on the maximum power that the customer used during the billing period[4].  Typically such demand charges will be based on the average consumption for some 15 minute period.

 

Cost has been a significant factor that limited the use of demand charges to commercial and industrial customers.  Demand meters are more costly to manufacture, in that they do more than just accumulate the amount of energy that goes through the meter.  Demand meters are more expensive to read, in that the meter reader has to note two quantities and has to manually reset the demand register.  These two cost factors are lesser issues in regard to determining residential demand now that the industry has moved significantly to Advanced Meter Reading (AMR) and to Advanced Meter Infrastructure (AMI[5]), both of which automatically collect consumption data, including for 15 minute intervals.

 

Historically residential demand charges was thought to produce an insignificant shift of revenue among residential customers.  The reasoning was that, though residential customers are different in size, they have a similar load pattern.  A customer using 1,000 KWH a month would have ten times the demand as a customer using 100 KWH a month.  Implementing a demand charge that collected an amount equal to 20% of the energy revenue collected from the larger customer would also collect an amount equal to 20% of the energy revenue collected from the smaller customer.  There would be no revenue shift among these residential customer, at least for consumption.  However, the utility would have had to install more expensive meters, which would have increased the monthly customer charge of both customers without providing a significant benefit to the utility or to the customers.

 

The move to AMR and AMI has reduced the cost of determining the demand for residential customers.  Now the cost of determining metered demand is not an issue in differentiating between customers with different consumption patterns.  Customers who should be paying a demand charge equal to 30% of their energy payments can be distinguished from customers who should be paying a demand charge that is only 10% of their energy payments.  Further, on site generation has changed the paradigm that residential customers have similar load patterns, so that now the industry knows that there are the 30% customers versus the 10% customers and can bill them appropriately.  Indeed, for houses with sufficient on-site generation, the revenue from the demand charge could be several times the revenue from the energy charge, especially when the energy charge vanishes for a net zero home.

The growth in AMR and AMI along with the growth in residential on-site generation makes this an appropriate time for restructuring residential tariffs to include a demand charge to collect the cost of the distribution utility owning the power lines.  The energy charge should continue to collect the cost of generation and transmission, though the energy charge should be time differentiated to reflect the real time value of generation and transmission, as well as the associated energy losses.



[1] The creation of Independent System Operators (ISOs) is alleged to have brought competition to the generation sector of the electric industry.  However, many distributed generators, such as roof top solar, do not experience the real time market prices set by their local ISO.  This distorts the market for distributed generation.

[2] The creation of ISOs is also alleged to have brought competition to the transmission market.  But ISOs compensate many transmission lines on a cost of service basis, through a monthly fee, though they charge geographically differentiated prices based on line losses and line congestion and generally don’t compensate for loop flow or parallel path flows, such as PJM imposes on TVA and on the Southern Company, both of which have lines in parallel to PJM>

[3] Telephone customers occasionally receive a business signal, indicating that the called party is using his/her phone.  More rarely, customers will receive a circuits business signal, indicating that intermediate wires are in full use, not that the called party is using his/her phone.

[4] Demand charges come in a variety of forms including contract demand, thermal demand, and ratcheted demands, a distinction beyond the scope of this discussion.

[5] AMI is generally distinguished from AMR in that AMI generally includes the ability to communicate both ways, from the meter to the utility and from the utility to the meter/customer location.  The ability to communicate from the utility to the meter allows the utility to control devices that the customer has opted to put under the utility’s control such as electric water heaters, air conditioning compressors, and swimming pool pumps and heaters.

Net Metering–Morphing Customers Who Self Generate

The U.S. Public Utilities Regulatory Policy Act of 1978 started a flood of non-utility generation, initially a few very large cogeneration plants and recently a large number of small roof top solar generation.[1]  The rapid growth in the number of small roof top solar generators requires the electric industry to develop a pricing plan that is fair to traditional customers as well as to the hybrid customers, those still connected to the grid but with some self generation.

Electric utilities support their pricing decisions with class cost of service studies  (CCOSS).  The CCOSS allocates the utility’s revenue requirement[2] to groups of customers, called classes.  Classes of customers are claimed to be homogeneous, such as being of a similar size, but more often as having similar load patterns.

Some costs in a CCOSS are allocated based on the number of customers, perhaps weighted by the cost of meters and services.  Fuel costs are allocated based on energy through the meter, though often weighted by the losses incurred to reach the  meter.  A large portion of the costs are allocated based on demand, the amount of energy used by the class during the times of maximum stress on the utility, or at times of maximum stress upon portions of the utility, such as on the generation, the transmission, distribution.  Utilities are concerned about recovering these demand related costs as customers morph from being a full requirements customer to being hybrid customers.

Electric utilities have long alleged that the homogeneity of residential load patterns allowed the utility to use energy meters, often called watt-hour meters, to determine how much each residential customer should pay each month.  The logic is that the allocation process brought costs into the rate class based on the customer’s demand.  Further, homogeneity means that the amount of energy through the meter is proportional to the customer’s demand.  The utility could collect roughly the right amount of money from each residential customer by charging residential customers based on their energy consumption[3] instead of charging residential customers based on the demand.

Charging customers based on energy allowed utilities to reduce substantially the cost of owning and reading meters without significantly distorting how the revenue to cost ratio from each customer.  At least until roof top solar substantially reduced the amount of energy that goes through the meter without necessarily reducing the customer demand.  Thus, with roof top solar, the revenue collected from the customer goes down greatly while the costs brought in by the customer demand goes down only slightly.

The growth in roof top solar coincides with the growth of Advanced Metering Infrastructure (AMI).  AMI often includes automatic meter reading and interval metering .  Automatic meter reading generally means replacing the person walking door to door with equipment.  The carrying cost of the equipment is often less than the cost of the human meter reader, allowing AMI to pay for itself.  Interval metering means collecting the amount of energy delivered during small time intervals, generally one hour (24×7), though sometimes on an intra-hour basis.  These interval readings are the demands in the CCOSS.

The intra-hour meter readings made possible by AMI would allow electric utilities to charge all residential customers based on their maximum demands, the determinant used in CCOSS to allocate costs to customer classes.  No longer would the utility have to rely on homogeneity assumptions in regard to residential customers.  The demand charge permitted by AMI would reduce the disparity between the lower revenue to cost ratio for residential customers with roof top solar relative to the revenue to cost ratio of standard residential customers.



[1] See

[2] the amount of money the utility needs to collect each year to continue functioning

[3] with a very inexpensive watt-hour meter

Net Metering–Reducing the Cross Subsidies

The U.S. Public Utilities Regulatory Policy Act of 1978 started a flood of non-utility generation, initially a few very large cogeneration plants and recently a large number of small roof top solar generation.  The large cogeneration plants sold power to utilities under individual contracts, such as those using the Ernst & Whinney Committed Unit Basis which I designed in 1984 for the Texas Study Group for Cogeneration and which was adopted that year by name by the Texas Public Utilities Commission.  The concept was adopted in other jurisdictions though generally without the explicit reference used by the Texas PUC.

The large number of roof top solar projects required a generic approach to pricing the output of non-utility generation.  A simple expedient has been net metering.  When there is a single meter on a customer premise, the meter can only measure the net amount into the premise.  Any generation just reduces the amount of electricity that the customer takes from the utility.  At some times, the generation will exceed the customer consumption resulting in an export of electricity to the utility.  Some jurisdictions extend the net metering concept to allow the exported energy to be offset against energy draws during other periods.

A problem with net metering is that the value of electricity changes rapidly, perhaps by a factor of 100 in the span of a few seconds.  Further, the change in value can be between positive and negative, as utilities are increasingly stressed by surpluses, especially at night and during periods of rapidly changing meteorological conditions.  This confounding situation makes net metering less and less applicable to the energy meter that has been a standard for measuring domestic consumption.  The advent of the smart grid and an automatic metering infrastructure may alleviate some of this issue, at least once the utility industry adopts real time pricing for retail consumption.

I mentioned the time span of a few seconds previously.  Independent System Operators create new economic dispatch order every 5 minutes, in that ISOs re-evaluate the relative merits of the available generation based on their relative fuel costs, as well as transmission constraints that can occur in less than a second, such as with a bolt of lightning.  The rapid change in the transmission system also occurs on the distribution system, which should similarly impact the price charged to retail consumers.  Decreasing the time span for pricing mixed deliveries of electricity should reduce the subsidies that can occur with net metering.