SunCommon of Vermont commissioned Synapse-Energy to perform a study that would show the benefits of before-the-meter (BTM) solar. The study shows there were almost $20 million of benefits for all of New England, about $1.3 million for Vermont during the very sunny period of July 1 to July 7; that period had very little wind, and thus almost no wind generation.
That joyous news was released with much fanfare, plus admonitions from SunCommon not to reduce solar subsidies, and to build more BTM solar systems to save the world. SunCommon is in the business of selling, financing, installing BTM solar systems.
Determining the Savings
Where do these savings come from?
ISO-NE, the grid operator, monitors the loads, MW, and local market prices, $/MWh (LMPs) of various parts of the NE grid, on a minute-by-minute basis.
LMPs are low at low loads, such as late at night, and high at high loads, such as on hot days in summer. S-E constructed a graph of historic LMPs versus historic loads using ISO-NE data. See page 18 of URL.
S-E claims BTM solar reduced NE loads, as seen by ISO-NE, and that lowered LMPs, which is true, hence the LMP savings.
S-E claims BTM solar reduced ISO-NE demand charges, which is true, hence the Capacity savings. See table on page 5 of URL.
The graph on page 8 of the S-E report shows the load, as seen by ISO-NE, is reduced by BTM solar (such as household rooftop, building roofs and other small projects on distribution grids) from the yellow line (what statistically the LMP would have been without BTM solar) to the blue line (what statistically the LMP would have been with BTM solar).
Solar is shown to be decreasing from its maximum around noontime (about the middle of the yellow band) to a low value in late afternoon/early evening when the peak demand for the day occurs (towards the end of the yellow band); solar does not reduce peak NE loads, as seen by ISO-NE.
The lower loads, as seen by ISO-NE, reduce LMPs, as shown by the graphs on pages 8 and 15.
NOTE: S-E failed to show the hours on the horizontal line of the graph, a major oversight to present a clearer picture. See page 8 of URL.
- Such savings occur only during the midday part of very sunny summer days.
- The savings are only of significance on very sunny days in summer.
- There would be very little such savings during winter.
NOTE: A more complete graphic is shown on this ISO-NE URL
NOTE: All NE grid peak loads in June and July occurred between 5 and 6 pm, when solar was already near minimal, i.e., solar was not much help reducing the peak demands. See Peak Load (Day & Time) graphic in this ISO-NE URL.
Brokers in the NE wholesale market buy at LMP prices, add a markup, and sell to utilities at higher prices (“what the market can bear”), similar to the bid and ask system on the stock market. When utilities are eager to buy, as on hot, sunny summer days, markups are higher, i.e., profits for brokers are higher. That means brokers pocket some of the claimed savings mentioned by S-E, instead passing them on to utilities. Infamous, greedy Enron comes to mind.
SE Method of Calculating Savings Due to Solar
S-E calculated saving = quantity of solar x LMP (without solar) - quantity of solar x LMP (with solar) for each minute during which solar was active.
S-E also calculated the ISO-NE imposed demand charge (load impact) for each minute during which solar was active.
The total savings are summarized in tables on pages 4 and 5.
Those savings should be reduced by the higher markups of the brokers.
S-E Savings May Be Overstated
- Almost all electricity is bought by utilities directly from producers under power purchase agreements, PPAs.
- That electricity is not traded on the LMP market and therefore not subject to any LMP variations.
- Only the percentage traded on the LMP market would have the savings mentioned by the S-E report.
- S-E should specifically mention the quantities of electricity traded on the LMP market and calculate the savings for only these quantities.
Savings to Ratepayers Likely were Minimal
From a VTDigger article:
"SunCommon Co-President James Moore said the report, commissioned by SunCommon and prepared by Boston-based Synapse Energy Economics, confirmed their suspicion that solar helps lower peak demand during heat waves."
To repeat, solar does not reduce peak demand because peak demands have shifted to late afternoon/early evening.
New rooftop solar installations have decreased in 2017 and 2018, so SunCommon is interested in boosting sales by touting solar savings.
Solar is a midday affair, but peak demands occur in late afternoon/early evening, when solar is minimal, according to ISO-NE real time demand data.
S-E Ignored the Huge Costs of Net-Metered Solar to Ratepayers
In 2016, GMP bought 71970 MWh, at a cost of $15,699,137, or 21.813 c/kWh. GMP paid to owners about 18 c/kWh and 3.813 c/kWh was GMP costs.
GMP had about 150 MW of BTM solar on its system in August 2018, which would have an annual production of about 150 x 8766 x 0.140 = 184,086 MWh. GMP would buy about 2/3rd of that, or 122,724 MWh. See table.
GMP could have bought that midday electricity at wholesale prices (LMPs plus markup) of about 6 - 7 c/kWh, on almost all midday times, except on hot summer days, when LMPs likely would be higher. See graph on 15 of URL.
Because net-metered is generated close to the user, there are other savings which amount to about 3 c/kWh, as estimated by GMP.
So the real value of solar to GMP is about 9 to 10 c/kWh, say 9.5 c/kWh.
That leaves a difference of about 11.793 to 12.793 c/kWh, which is rolled into rate schedules and which every ratepayer gets to pay one way or another.
The more net-metered solar (to the happiness of SunCommon), the more other ratepayers pay.
The additional cost paid by ratepayers, due to additional net-metered solar, would far exceed any S-E calculated savings, which primarily occur only on sunny summer days.
S-E ignored the extra cost to ratepayers of about $15.1 million/y, which far exceeds the estimated savings of $1.3 million of the S-E report.
S-E also ignored the cost of various subsidies, which make solar appear at least 30 to 40 percent less costly/kWh than without subsidies.
GMP cost, $/y
BTM solar production
150 x 8766 x 0.140
184086 x 1/3
Bought by GMP
184086 x 1/3
GMP program costs
184086 x 2/3 x 3.793
GMP real value
Extra cost to ratepayers
S-E Ignored Cost Impact on Other Generators
Also, the more solar, the more other generators have to reduce their outputs.
Such large output reduction in a few hours is less efficient (more Btu/kWh, more CO2/kWh), than steady operation.
That inefficiency works its way into upward LMPs.
As midday solar increases, so-called “Duck Curves” occur.
The increased up/down ramping of generators, the more frequent starting and stopping of generators, and the increased synchronous generators on standby to deal with Duck Curves, cause the generators to be less efficient and wear out more quickly (just as a car), as proven in Southern Germany and California, both areas have high levels of solar.
S-E Study Evaluation
The S-E study appears to be not a balanced presentation of the real situation. In fact, its likely impact may be a deception perpetrated on an unsuspecting, unknowing, ignorant, naïve, lay public. The SE study deserves a special place....
Significant subsidies that have contributed to the increase in Vermont solar are being reduced. Vermont has seen a decrease in solar installations in 2017 and 2018.
Changes to the state’s net metering rules and a 30 percent federal tariff on imported solar panels have contributed to that decrease, according to a recent report from the state’s Clean Energy Development Fund.
GMP has been one of the players advocating for a decline in the above-market rate paid to net-metered customers who feed excess electricity onto the grid.
Josh Castonguay, chief innovation officer of GMP, contends increased energy storage, not more solar installations, would reduce future peak demands and lower electricity costs. GMP saved $600,000 during this year’s August 6 regional peak by drawing on energy stored in batteries, like the Tesla powerwall 2.0, he said. GMP had to make investments in solar and battery systems totaling at least $20 to $25 million to achieve the cost reductions.
“Initial deployment of solar several years ago did a good job of providing value to the peaks and energy up to a point – the largest benefits came early on,” he said. “We now have a shift with peaks pushed to late afternoon/early evening so the solar (being minimal in late afternoon/early evening) is less cost effective for customers to reduce peak demands.”
High Electricity Prices for RE in New England: The highly subsidized wholesale prices of wind and solar paid by utilities to producers are much higher than in the rest of the US, because of New England’s mediocre wind and solar conditions.
Wind and Solar Far From Competitive with Fossil in New England: The Conservation Law Foundation claims renewables are competitive with fossil. Nothing could be further from the truth. Here is a list of NE wholesale prices and Power Purchase Agreement, PPA, prices.
NE field-mounted solar is 12 c/kWh; competitively bid
NE rooftop solar is 18 c/kWh, net-metered; GMP adds costs of 3.813 c/kWh, for a total of 21.813 c/kWh
NE wind offshore until recently about 18 c/kWh
NE wind ridgeline is at least 9 c/kWh
DOMESTIC pipeline gas is 5 c/kWh
Russian and Middle East imported LNG is at least 9 c/kWh
NE nuclear is 4.5 c/kWh
NE hydro is 4 c/kWh; about 10 c/kWh, if Standard Offer in Vermont.
Hydro-Quebec imported hydro is 6 - 7 c/kWh; GMP paid 5.549 c/kWh in 2016, under a recent 20-y contract.
NE annual average wholesale price about 5 c/kWh, unchanged since 2009, courtesy of low-cost gas and nuclear.
NOTE: Vineyard Wind, 800 MW, fifteen miles south of Martha’s Vineyard, using 8 or 10 MW turbines, 750 ft tall.
Phase 1 on line in 2021, electricity offered at an average of 8.9 c/kWh over 20 years
Phase 2 offered at an average of 7.9 c/kWh over 20 years
NOTE: The NE grid is divided in regions, each with Local Market Prices, LMPs, which vary from 2.5 - 3.5 c/kWh from 10 pm to about 6 pm; slowly increase to about 6 - 7 c/kWh around noon time, when solar is maximal; are about 7 - 8 c/kWh in late afternoon/early evening (peak demand hours), when solar is minimal. Unusual circumstances, such as power plant or transmission line outages, can cause LMPs to increase to 20 - 40 c/kWh, and even higher when such events occur during peak demand hours.
NOTE: The above prices would be about 50% higher without the subsidies and even higher without cost shifting. See Appendix.
NOTE: Here is an ISO-NE graph, which shows for very few hours during a 13-y period were wholesale prices higher than 6 c/kWh. Those prices are low because of low-cost gas, low-cost nuclear and low-cost hydro. The last four peaks were due to:
- Pipeline constraints, aggravated by the misguided recalcitrance of pro-RE Governors of NY and MA
- Pre-mature closings of coal and nuclear plants
- Lack of more robust connections to nearby grids, such as New York and Canada. See URLs.
Wind and Solar Conditions in New England: New England has highly variable weather and low-medium quality wind and solar conditions. See NREL wind map and NREL solar map.
- Wind electricity is zero about 30% of the hours of the year (it takes a wind speed of about 7 mph to start the rotors)
- Wind is minimal most early mornings and most late afternoons/early evenings (peak demand hours), especially during summer
- Wind often is minimal 5 - 7 days in a row in summer and winter, as proven by ISO-NE real-time generation data.
- About 60% is generated at night, when demand is much less than during the late afternoons/early evenings
- About 60% is generated in winter.
- During winter, the best wind month is up to 2.5 times the worst summer month
- New England has the lowest capacity factor (about 0.262) of any US region, except the US South. See URL.
- Solar electricity is strictly a midday affair.
- It is zero about 65% of the hours of the year, mostly at night.
- It often is minimal 5 - 7 days in a row in summer and in winter, as proven by ISO-NE real-time generation data.
- It is minimal early mornings and late afternoons/early evenings
- It is minimal much of the winter months
- It is minimal for several days with snow and ice on most of the panels.
- It varies with variable cloudiness, which would excessively disturb distribution grids with many solar systems, as happens in southern California and southern Germany on a daily basis. Utilities use batteries to stabilize their grids.
- During summer, the best solar month is up to 4 times the worst winter month; that ratio is 6 in Germany.
- New England has the lowest capacity factor (about 0.145, under ideal conditions) of any region in the US, except some parts of the US Northwest.
NOTE: Even if the NE grid had large capacity connections with Canada and New York, any major NE wind lull and any major NE snowfall likely would affect the entire US northeast, i.e., relying on neighboring grids to "help-out" likely would not be prudent strategy.
Wind Plus Solar:
- Wind plus solar production could be minimal for 5 - 7 days in summer and in winter, with snow and ice on most of the panels, as frequently happens during December, January and February, as proven by ISO-NE real-time generation data.
If we were to rely on wind and solar for most of our electricity, massive energy storage systems (a few hundred GWh-scale for Vermont, multiple TWh-scale for NE) would be required to cover multi-day wind lulls, multi-day overcast/snowy periods, and seasonal variations. See URLs.