By Rud Istvan,
WUWT readers emailed Charles the Moderator concerning two flurries of renewable Fake News PR in the past week:
1. The breathlessly reported planned FPL addition of the world’s largest storage battery for Florida’s Manatee County solar farm.
2. The less reported Bloomberg New Energy report that batteries were now becoming competitive in some grid applications/locations thanks to steeply declining costs.
ctm asked me to take a crack at an article covering this Fake News. This guest post tackles these items in sequence (and they are related). The research was somewhat complex, so rather than provide numerous links, sufficient pointers are provided that will take any interested readers to the relevant underlying documents via simple Google searches. Else, just enjoy the deceptive saga.
FPL PR
Many news outlets (Ars Technica, Forbes, every MSM paper in Florida) reported on the Florida Power and Light (FPL) announcement last week about “the world’s largest solar powered battery system—4x the next largest” (at Hornsdale, South Australia). Paraphrasing for brevity, it will be built on the 40 acre site of soon to be demolished almost 50 year old twin oil/natural gas-steam generating units in Parish, Florida. It will have 409MW and be able to produce 900MWh of energy from FPL’s adjacent Manatee solar farm and another (of equal size) to be built nearby. It will provide clean, cost effective electricity.
NOTHING about this FPL summary is correct except the location.
Battery capacity is not specified in MW, rather in MWh. Cost is ~ MW. Hornsdale is 100MW and only 129MWh. So the Manatee battery is actually (900/129) seven times more capacity. Oops, an ignorant PR under brag.
It does not replace the old oil converted to gas fired steam generating units, unlike what FPL implied. As old and decrepit as they are, their current August peak capacity (all that matters in Florida) is still 1618MW. (This and following data are from FPL’s “10 Year Power Plant Site Plan 2018-2027”, submitted to FloridaPSC.com and running about 270 pages of (1) capacity, (2) demand, and (3) capacity addition rate planning. Parish’ present 1618MW is in Part 1.) Parish baseload generating capacity is being replaced concurrently by 1778MW of new CCGT at Lake Okeechobee.
It is not powered by solar. The nameplate capacity of the Manatee solar farm and the nearby one now under construction is 74.5MW each. Per Part 1 of the FPL 2018 Plan, the August peak solar capacity factor is 52%. Both solar farms can provide (unless/until a hurricane) (974.5*2*0.52) 77.5MW, just 19% (less in winter) of what the battery needs. The rest will come from FPL grid CCGT additions.
It is NOT cheap. FPL refuses to say what it will cost. (Hey, I really tried to find out.) BUT, there is a small pilot plant equivalent (10 MW) that was installed at FPL’s Babcock Ranch solar farm (also 74.5MW nameplate) in 2018. From the FPL supplied pictures (Babcock Ranch actual, Parish FPL conceptual), the batteries are the present utility industry grid standard, high temperature (350C) Sodium Sulfur, with individual 1 MW modules looking like 20 foot shipping containers stood on end. That 10 MW/40MWh cost $15 million (digging through FPL and local press at the time)–and like Parish will be–is used for peak load. (Note: this is about industry standard capacity cost for that type battery. I covered two PG&E sodium/sulfur grid installations from 2010 and 2014, with illustrations, in essay California Dreaming in ebook Blowing Smoke.) So the Parish battery will cost ([409/10]*15) about $615 million capital dollars. By comparison, according to EIA Nov 2016, equivalent gas peaker turbine capacity would cost $450 million. (Remember this battery is 80% energized by grid generation burning natural gas, just like peaker turbines.) A bad economic deal for Floridians, if good for FPL green PR.
And finally, from the 2018 Plan part 3 rate considerations, we learn that this battery is being financed under SoBRA rate rules. Florida PSC says SoBRA guarantees FPL the full required revenue/cost recovery (plus return) via increased utility rates for ‘green’ solar, battery, and related activities. FPL’s claim that this is ‘clean and cost effective’ is only half correct. Maybe.
Bloomberg New Energy Forum
This ‘research’ unit of Bloomberg has been churning out rosy renewable projections for years. Following is their April Fools 2019 gem.
For starters, BloombergNEF believes the nonsense put out by the EIA concerning the levelized cost of renewable energy (LCOE, an annuitized version of net present value). In a guest post over at Judith Curry’s Climate Etc, Planning Engineer (very senior exec at one of the US’ largest electrical utilities) and I wrote a deconstruction of EIA’s 2015 onshore US wind LCOE titled ‘True Cost of Wind’. Easily found using the search tool on her site. We corrected all the EIA ‘cheats’, and used the Texas ERCOT grid (~10% wind penetration) for cost of backup and incremental transmission. Unlike the BloombergNEF figure of ~$95/MWh for 2015, the true cost of onshore wind is about $144/MWh. So you already know this new table is a renewable April Fools joke.
There are three recent reference points for grid battery storage compared to the chart’s purplish/pink battery line.
First is Tesla’s LiIon installation at Hornsdale, South Australia. That came on line November 2017 after the infamous renewables induced SA blackout of 2016. 129MWh cost about $50 million (Australian, as reported in Australia). The current exchange rate is $1Aus = $0.71US. So Hornsdale cost about $35.5 million US, or about (35500k/129) $275k per MWh. The LCOE would be higher, because the capital is up front and the ‘fuel savings’ are annuitized. Bloomberg is ‘only’ two plus orders of magnitude off. But Hornsdale was a rushed Elon Musk PR one off, so maybe not a fair comparison.
Second is the Tesla PowerWall v2. Excluding the power electronics (which do scale some), it presently costs $6700 for 13.5 KWh and benefits from mass production in the GigaFactory. That math works out to $496/MWh, again higher when annuitized into LCOE. BloombergNEF is ONLY off by 1.65x.
Finally we have a direct grid comparison using FPL’s 2018 Babcock Ranch, which was 10MW for 40 MWh, exactly the BloombergNEF 4 hour standard. $15 million is (15000k/40MWh) $375k/MWh—ONLY off by ONE order of magnitude. Order of magnitude errors are not forgivable. They are Fake News.
Thus demonstrating that renewable Fake News remains alive and ‘well’.
H/t to WUWT commenter RK and ctm for the inspiration to get going again.
I think Rud will remember the debacle of Progress Energy being allowed to collect pre-constuction costs for a nuclear plant without any guarantee that it would be built. Guess what? People paid, plant later cancelled, no refunds allowed.
Why should we believe any future plans promoted as good for us without monetary guarantees back to the customer.
That sounds like 50% of the national construction projects (housing, highways) here in Socialist South Africa!
Rud,
What happens when one of those vertical 350 C Sulfur batteries tips over/breaks open in a hurricane? I would think explosion that would cause chain reaction with nearby batteries, and a full scale disaster. Or are they proof to that kind of damage?
My opinion only.
These modules are bolted onto concrete pads. Toppling is not the hurricane issue.
If they lose power (e.g. faulty transmission line power) they ‘freeze’. Whether they can be slowly restarted from ‘freeze’ depends. Is very complicated..
The prevailing notion with the media/political class infecting the general public is the false assumption that Moore’s Law applies, the exponential development of computer chip technology, is an analogue for battery development; that a future world powered by ‘clean’ energy and batteries is ‘just around the corner’ and anyone who doesn’t get with the program is a Luddite.
https://www.cfr.org/blog/why-moores-law-doesnt-apply-clean-technologies
How clean and cost effective will these batteries be for the next generation, after they don’t work anymore and corrosion takes its course? Seems to me around 1990 or so, corporations abandoned long range thinking and adopted myopic, immediate gratification to boost their images or profits.
Thanks for the analysis. I have a question.
“August peak solar capacity factor is 52%” – How does that work? Capacity factor equals actual output over nameplate capacity potential output. I have been trying to figure out how to calculate utility-scale solar farm output and it is not as simple as I had hoped. The nameplate capacity at the solar farm is 74.5 MW but at what solar insolation value? I dug around and decided to use 800 waters per square meter as the nameplate capacity rate. If I am correct or at least semi-close, then when insolation is greater than that the actual output is greater than the nameplate. In that case then 52% capacity factor is not unexpected. Am I on the right track?
My starry-eyed anti-nuclear relatives in the San Francisco bay area tell me California doesn’t need to keep Diablo Canyon on its power grid anymore, because wind and solar backed by battery storage and by pumped hydro can do it all.
And for less cost, too. (Or so they claim.)
When I express my skepticism as to whether wind and solar backed by grid-scale battery storage and by pumped hydro will ever work, the knee-jerk response from my California relatives is that the cost of these technologies is coming down so quickly that we really don’t need to worry about the future economics of the renewables.
My admittedly complicated reply to their unbridled faith in the renewables goes like this:
I explain that my skepticism of these claims is based on the complete lack of realistic cost and schedule analysis studies for large-geography wind and solar proposals, studies that use commonly accepted methods for doing engineering feasibility estimating.
As an illustration of what I’m talking about, I offer them an example of how one should go about costing a mostly renewable power grid that covers a large geographic area.
Assume that California and Nevada work together to create an 80% renewable electric grid for their two states. The grid will be developed and managed by a quasi-government corporation, California-Nevada Clean Energy Corporation, ‘CalVada CEC’ for short.
Through legislation and agreement, California and Nevada jointly grant CalVada CEC carte blanche authority for overcoming any technical, financial, or regulatory obstacles that might get in the way of building and deploying this 80% renewable joint power grid.
As part of a major analysis study done for future regional planning, a feasibility design the CalVada CEC grid is developed to a level of engineering detail which can support a reasonably tight feasibility cost and schedule estimate for what it will take to build and operate this new system for the next fifty years.
Assume a set of grid reliability requirements equivalent to what is now in force today. Assume that the land needed for locating the solar arrays, the wind farms, and the battery storage facilities will be allocated through a process of eminent domain, with fair prices and rents being paid to the land’s current owners. In addition, assume that other windmills will be located off California’s coast wherever they are best placed to maximize wind capacity factors.
Locate the pumped hydro facilities wherever they are best placed to support the integrated grid design, again using a process of eminent domain. Fast track the environmental review processes for all elements of the renewable energy grid system so that regulatory oversight of CalVada CEC’s energy facility siting, construction, and plant operations is minimized to the maximum extent possible.
The CalVada CEC’s grid feasibility design must include specific engineering details for the particular solar arrays to be used and their proposed locations, the particular wind mills to be used and their proposed locations, the particular battery and pumped storage facilities and their locations, the routes and configurations of the power transmission corridors, and the configurations and locations of the power distribution and control facilities.
After the feasibility engineering design for the CalVada Clean Energy Grid is complete, and it is available in enough level of detail to reliably identify what each major phase and sub-phase of the project will cost, then analyze the cost and schedule risks of the project as a whole in order to determine what range of total costs can be expected, and how long it will take before the 80% figure for both states can be finally achieved.
So concludes my response to my anti-nuclear San Francisco bay area relatives. What kind of reaction do I get from them, once the glaze clears from their eyes?
Mostly their reaction is delivered as emotionally-charged pushback citing high level studies previously done by renewable energy experts — studies which claim to prove that the future costs of wind and solar backed by energy storage will be less than what nuclear power delivers today.
The scary part is that if you went to a legislator who was sponsoring legislation mandating the 80% renewable grid the response would be “emotionally-charged pushback citing high level studies previously done by renewable energy experts”. We don’t need to do no stinking analyses we know the answer. It cannot end well.
Given the South Australia battey is quoted in the above article, it might be opportune to mention a major (A$650M) solar project at Port Augusta, in South Australia has been scrapped according to a report on the Australian Broadcasting Commission news page.
https://www.abc.net.au/news/2019-04-05/solar-thermal-plant-will-not-go-ahead-in-port-augusta/10973948
Key points:
The plant was announced by the former state government in 2017
It received a Federal Government loan and was promised a State Government electricity contract
SolarReserve says it cannot secure funding for the project
No funding despite government guaranteed loans and Government guaranteed power supply contracts. What can possibly go wrong?
Excellent news! I’m not sure which I hate more, the the bird birmers or the bird choppers.
All this death and destruction caused by windmills and solar thermal, is completely unnecessary.
This has been kept very quiet in the Aussie MSM..
Wasn’t it meant to power the Whyalla steel works and enrich the Indian tycoon Gupta who bought them?
Ah, yes. I suppose it was intended to replace the coal power plant in Port Augusta that was rased in a fit of virtue signalling and to avoid the temptation of going back to coal no matter how dire the grid circumstances get.
This was a detailed and accurate assessment of utility-scale renewable projects, but one sentence should be highlighted:
“Florida PSC says SoBRA guarantees FPL the full required revenue/cost recovery (plus return) via increased utility rates for ‘green’ solar, battery, and related activities.”
The state utility commission isn’t just “saying” this guarantee on recovery; it is AUTHORIZING it. The primary driver behind utility-scale renewable projects is not the utilities, although some like Green Mountain Power in Vermont actually have sipped the Flavor-aid. Rather, the driver is the collective, state utility commissions.
There was a time when the investor-owned utilities (IOUs) were primarily concerned with power quality, system reliability, and grid stability. These factors necessitated a practical and prudent approach, knowing that the respective, state utility commissions and/or consumer advocates would scrutinize the costs thoroughly. However and at the continued direction of the commissions, IOUs over the past decade have abandoned this approach. Why? Commissioners, who are political appointees of the governor, compose the commissions. As political appointees in the Age of Renewables, they are implementing a political ideology rather than a proven practicality.
Even in states where IOUs are not permitted to own “traditional” power generators (e.g., Massachusetts) commissions have not only granted IOUs waivers to install utility-scale renewable “generation,” they have also mandated their installation. What better way to hasten a societal change of utmost importance (less than 12 years remaining until civilization’s demise, you know) then through (mostly) voiceless dupes. The public perception is that the “greedy” IOUs are the promulgators behind such silly projects, when in reality, the commissions are driving the bus. The IOUs accept this approach because (1) they are guaranteed recovery (which protects the shareholders’ investment), and (2) their cooperation with the commissions – now provides for a smoother rate case (separate of renewable portfolio supply charges) – then.
The consumer advocates may protest the resulting increases, but even they are compelled to accept (to a greater extent than other issues like reliability) the “rightness” of utility-scale renewable projects. Although IOUs know renewables are not the solution to the demonized fossil fuels, the position of many within the industry is that the state regulatory choo-choos have already left the stations. Those still trying to assert the practicality approach are viewed increasingly as cranks.
Thanks for that post, Tom.
Our local politicians seem to be as big a part of the problem as the national politicians. Maybe more so, as the locals fly “under the radar” for the most part, and people don’t know about their crazy plans, such as promoting windmills, or “sustainability”, until they are already implimented and the damage is done.
More scrutiny on the local level is required.
I found an interesting post with lots of detail and commentary:
https://hanshyde.com/2019/03/30/solar-storage-closing-gas-plants/
My quick Google search indicates Parrish, Florida is about 40 feet above sea level.
So, flood insurance shouldn’t be a problem ??
How do we counter this?
https://reneweconomy.com.au/the-stunning-numbers-behind-success-of-tesla-big-battery-63917/
A major chunk of the earnings occur when grid prices soar towards the legal limit of A$14,000/MWh. That happens because the grid has inadequate dispatchable capacity, so is designed to create those conditions that give the battery a margin. Even so, its contribution is derisory, and failed to prevent rotating power cuts last summer.
Batteries are always a problem with a limited life. One would be better off with a gas-fired generator attached to one’s house for when the whole grid fails. The way that Warmista control is gaining ground, that could be tomorrow.
” costs $6700 for 13.5 KWh and benefits from mass production in the GigaFactory. That math works out to $496/MWh”
oops
you mean $496/KWh or $496,000/KWh
you mean $496/KWh or $496,000/MWh of course.
Grid scale battery storage would materialize on its own if power companies offered power metered at the current cost of power which would be transmitted over the power lines and a smart controller in the home could be used to charge batteries or reduce grid draw and increase draw from battery according to price at the moment. The only thing that would be required would be for the power utility to transmit the current price of power. In times when there is an abundance of power and they are paying other utilities to take it (grid operator price goes negative) such as during a burst of wind power at 2am, homeowners would be able to get free power to charge their batteries for use when the surplus condition passes. It would use market conditions, nobody would force anyone to take such a metering plan but people could save considerable money if they did and would be able to shift their load on the grid from high demand periods to periods of low demand. All it takes is a smart charging controller that is able to sense the current price of grid power and decide how much to draw from the grid and how much to draw down or charge up local batteries. If adopted on a wide scale, the utility becomes more stable as demand begins to follow supply in real time and the consumer saves money.
“All it takes is a smart charging controller that is able to sense the current price of grid power and decide how much to draw from the grid and how much to draw down or charge up local batteries. If adopted on a wide scale….?’
https://www.manhattan-institute.org/green-energy-revolution-near-impossible
A grid consisting of solar wind and batteries for a couple of days support PLUS replacing all the ICEs you see about you on the roads today? I get the bit about price being a better market signal than rationing BUT….!!
There are several schemes in different countries where the grid has first call on the use of domestic battery installations, in exchange for some reduction in installation cost for the household to cover the reduced flexibility they have to use the capacity themselves. Grid usage is mainly for stabilisation, which requires the batteries to operate part charged so they can charge or discharge to help balance the grid. These are poor deals for consumers, who end up subsidising the operation. The problem is that batteries are not cost effective.
The Australian April 5 2019 reports-
“A $650 million renewable energy project brokered by South Australia’s former Weatherill government has been abandoned because the US company behind the project is unable to secure finance.
SolarReserve’s planned 150 megawatt solar thermal station near Port Augusta, 310km north of Adelaide, was hailed by Labor as a game-changer when it announced in August 2017 the company had won the contract to supply the state government’s energy needs for 20 years.
To be propped up by $110 million of federal funding that was negotiated by Nick Xenophon, the project was to create 650 jobs during construction and be operational by next year…..
The plant would have seen concentrated solar thermal technology harness the sun’s power to generate electricity by using lenses and reflectors to concentrate sunlight, heating molten salt to produce steam to drive a turbine.
It was expected fill a void in Port Augusta after its coal-fired Northern power station closed in 2016, largely because of the rapid rise of renewables in South Australia.”
So the smart money has chickened out of solar and molten salt even with the usual handout and guaranteed public sector demand. At least our dilly Gummints are beginning to require dispatchable electrons from these subsidy miners and they’re not so keen anymore that they can’t engage in their old dumping practices for hydrocarbons to pick up the insurance tab and hide the real social cost in our power bills.
The portended retirement of Liddell coal station is beginning to flush out the fairy dust merchants and how many of Tesla’s cutting edge 2170 lithium lego bricks will be required to sort out the existing electron dumpers the punters were told would be free from Gaia?
https://anero.id/energy/wind-energy/2019/march
Just needs a little bit of fine tuning folks-
https://www.abc.net.au/news/2019-01-07/blackout-compensation-scheme-lowered/10692296
“I guess every dollar helps in this climate, and this is what we as a commission can control,”
“We’re doing our bit.”
LOL
Batteries pay for themselves.
The Second largest Grid scale battery in South Australia will be fully paid off in 3 to 4 years based on the first 18 months of operation. It has also saved the State Government about 12 million dollars by damping down spot prices.
South Australian Grid battery pays off 1/3 of initial investment in first year.
https://www.theguardian.com/technology/2018/sep/27/south-australias-tesla-battery-on-track-to-make-back-a-third-of-cost-in-a-year
South Australian Grid scale battery has protected taxpayers from artificially created spot prices saving 20 million dollars in the first year.
https://www.afr.com/business/energy/solar-energy/south-australias-big-battery-slashes-40m-from-grid-control-costs-in-first-year-20181205-h18ql1
“South Australian Grid scale battery has protected taxpayers from artificially created spot prices saving 20 million dollars in the first year.”
There is nothing artificial about these spot prices spikes.
No Michael, they haven’t paid for themselves. They have just lowered some of the spot prices from exorbitant to crippling. Others who know a lot more about energy than journalists are not that impressed. http://www.wattclarity.com.au/articles/2019/03/price-volatility-in-vic-and-sa-on-friday-1st-march/
http://www.wattclarity.com.au/articles/2018/12/dispatch-price-spikes-above-12000-mwh-in-vic-and-sa-today/
Wait till the next time the link to Victoria trips out under high inflow conditions, then see how good the batteries are.
If the SA government hadn’t closed the coal fired power station, the SA consumer would have had a lot cheaper power and there would be no need for the batteries.
The whole point of the SA battery is to take the load when the Victoria grid link trips or SA power plant goes offline (apart from providing cheaper peak electricity and smoothing transition as wind or solar ramps down as predicted during the day).
A grid scale battery responds faster than running reserve.
It needs to respond faster Griff because it doesn’t provide any inertia. It also is effectively only 20MW input (go back and look at the contract) so when the Victoria line carrying 700MW trips out on overload, the battery provides basically support. And the short circuit levels from an asynchronous system lower the system stability even more. But that, like most engineering, is beyond your comprehension.
Nonsense. The Heywood interconnector is 800MW. Murraylink another 200MW. The battery is discharged at only 30MW except for extremely rare occasions when it has briefly used its 100MW maximum discharge capacity.
It Doesn’t
There is a time/ load rating in the protection systems on the components in the grid. The rated values for MCR are significantly lower than what fault loads just before trip can be. My understanding is Heywood is still only rated at 600MW west.
https://www.aemo.com.au/-/media/Files/Electricity/NEM/Security_and_Reliability/Congestion-Information/2017/Interconnector-Capabilities.pdf
The Murraylink is DC so its response to fault conditions is in the battery class.
And I left a word out of my post above that totally changed the meaning. When Heywoods trips on overload, the battery provides basically NO support.
Because lack of inertia was the nail in the coffin of the SA blackout, the grid operator changed the rules in SA so they have to have a certain number of GTs running and they will dispatch wind/ solar off to meet that target. They are now realising just how bad the unreliables are and why big thermal plant is what is needed.
Using expensive lithium batteries as your price reference is misleading.
EOS has been selling grid-scale batteries for $160/KWh for 2 years. $95/KWh if you’re willing to wait until 2022 for delivery.
https://eosenergystorage.com/products-technology/
https://www.businesswire.com/news/home/20170418005284/en/Eos-Energy-Storage-Orders-95kWh-Eos-Aurora®
Rated for 5,000 cycles, so that’s $0.032/KWh delivered. Or $0.019/KWh in 2022.
Normally you have to double the price for inverters to tie it into the grid, but when sited next to a solar farm you can share the inverters.
I forgot to say 6 months ago they announced a company with manufacturing expertise invested in EOS:
https://www.marketwatch.com/press-release/eos-energy-storage-secures-strategic-investment-from-and-manufacturing-partnership-with-holtec-international-2018-09-24
I believe this is the largest grid scale battery…
https://www.utilities-me.com/news/12329-abu-dhabi-launches-worlds-largest-virtual-battery-plant
Thanks to griff for the link, but…
What is a virtual battery?
and
“… Virtual Battery Plant with a capacity of 108 megawatts …”: Is it MWseconds, MWhours or MWdays.
Have none of these journalists been to any elementary school physics classes?
It is actually ten separate batteries in different locations, but by claiming them to be a single large “virtual battery” they get more publicity and fool simple souls like griff.
This is definitely a concept with possibilities, for example a hunter might shoot ten stray cats and turn them into a “virtual cougar”
A virtual battery is, as TTY points out, a large number of batteries connected distributed across a wide area. Expand on this further by grid-connected solar and battery systems, like the Tesla powerwall, via a smart meter and the “grid supplier” can take power from your domestic system, without your control, and supply the grid as required. The ultimate goal, as see it, is to move from a centralised system of generation to a distributed, grid-connected (Because we already have a grid) system of generation.
That is the myth. The reality is that whenever the weather is unfavourable, power may have to be wheeled in over great distances if meantime you shut the local power stations designed to support major city centres of demand. You may also need long and expensive connections to offshore wind farms. And you will need another beefing up of the grid in order to try to disperse power surpluses when the weather conspires to create them.
The reality is that batteries can only offer a very limited and expensive partial solution to these problems.
Rud: Thanks for the post. I understand the difference between MW and MWh, but am confused by part of your article:
“It [presumably the battery] will have 409MW and be able to produce 900MWh of energy from FPL’s adjacent Manatee solar farm and another (of equal size) to be built nearby.”
“[The battery] is not powered by solar. The nameplate capacity of the Manatee solar farm and the nearby one now under construction is 74.5MW each. Per Part 1 of the FPL 2018 Plan, the August peak solar capacity factor is 52%. Both solar farms can provide (unless/until a hurricane) (974.5*2*0.52) 77.5MW, just 19% (less in winter) of what the battery needs. The rest will come from FPL grid CCGT additions.”
It isn’t obvious to me why we want to compare the maximum output of the solar farm to the output of the battery (both measured in MW). I presume the grid will charge the battery whenever power is cheap. The nearby solar farm can charge by transmission over the shortest distance with the least loss of power, but I presume that transmission and generation throughout the grid will be optimized as a whole. When demand is high, solar power will be directed towards customers, not the battery. The more important performance parameter appears to be capacity (MWh) rather than output (MW).
The cost of a battery can be recouped by storing power when the price is low (because supply is high and demand is low) and then selling power for more when demand is high. The battery’s competition is a peaker plant that has both capital and fuel charges. The “fuel” charge for the battery is the cost of the power needed to charge it, the losses in the charge/discharge cycle. I’m not sure you analyzed finances from this point of view.
The only even remotely feasible “grid scale battery” is a pumped storage system. This, in turn, requires suitable geography – and enough water (which people forget about). During the last power shortage run in South Africa, it was also during one of its regular drought periods, so the two Drakensberg pumped storage systems couldn’t provide power, because there wasn’t enough water flowing into the lower dam to be pumped into the upper!
Rud,
You are quite correct to specify batteries as 100 MW/129 MWh. One without the other makes no sense.
Prices of batteries is one thing, turnkey capital cost is another.
Musk had to airship to Australia the entire Tesla supply by plane to make tight schedules.
Various capital costs were bandied about.
The Tesla Powerpack 2 system in Australia, the largest in the world, has a rated capacity of 100 MW/129 MWh (delivered as AC should be added, if that is known). The battery system feeds AC electricity to the high voltage grid or absorbs electricity from the grid to:
– Smoothen the variable output of the nearby 315 MW French-owned wind turbine system to minimize it “upsetting” the high voltage grid. Synchronous-condenser systems perform a similar function.
– Prevent load-shedding blackouts and
– Provide stability to the grid, during times other generators are started in the event of sudden drops in wind or other network issues.
In 2017, the TURNKEY capital cost was about US $66 million, or 66 million/129,000 = $512/kWh; this is a low price, because Tesla was eager to obtain the contract. Here is an aerial photo of the system on a 10-acre site.
https://www.mercurynews.com/2017/12/26/teslas-enormous-battery-in-australia-just-weeks-old-is-already-responding-to-outages-in-record-time/
https://en.wikipedia.org/wiki/Hornsdale_Wind_Farm
The FP&L system will be 409 MW/900 MWh, i.e., 900/129 = 6.98 larger than Hornsdale (delivered as AC should be added, if that is known).
In 2020, the TURNKEY capital cost of the batteries will be about 900,000 kWh x $450/kWh = $360 million, to be amortized over 15 years.
In 2020, the turnkey cost of the new CCGT plants will be about 1778 MW x $1.25 million/MW = $2223 million, to be amortized over 35 to 40 years.
In 2020, the turnkey cost of the new solar plants will be about 74.5 MW x 2 x $2.5 million/MW = $372.5 million, to be amortized over 25 years.
“It will have 409 MW and be able to produce 900 MWh of energy from FPL’s adjacent Manatee solar farm and another (of equal size) to be built nearby. It will provide clean, cost effective electricity.”
1) The battery will NOT be able to PRODUCE anything. It only charges, stores and discharges.
2) Any electricity passing through storage has about a 20% loss, on a high voltage AC to high voltage AC basis, i.e., it CONSUMES a lot of electricity.
3) Parish’s existing, 50-y old, inefficient plants, 1618 MW, will be demolished. On the cleared site will be the new batteries.
4) New, 55%-efficient, gas-fired CCGT plants, 1778 MW, will be built at Lake Okeechobee. THEY WILL REPLACE THE DEMOLISHED PLANTS.
5) The Manatee solar farm, and the nearby one, now under construction, are 74.5 MW each, area required about 1043 acres, at 7 acre/MW.
6) Production of both solar plants = 74.5 MW x 2 x 8766 h/y x 0.19, capacity factor = 248,165 MWh per YEAR.
7) The output of the solar plants will be minimal at 8 AM, maximal at noon-time, and minimal about 5 PM.
8) During mid-day, electricity is fed into the battery at the ALLOWED battery feed in rate, and no higher.
9) During evening, electricity is discharged from the battery at the ALLOWED discharge rate and no higher.
Below are some revisions and additions
Revisions:
In 2017, the TURNKEY capital cost was about 56 million Euros, about US$ 66 million, or 66 million/129,000 = $512/kWh; this is a low price, because Tesla was eager to obtain the contract. Here is an aerial photo of the system on a 10-acre site.
https://www.mercurynews.com/2017/12/26/teslas-enormous-battery-in-a…
https://en.wikipedia.org/wiki/Hornsdale_Wind_Farm
https://reneweconomy.com.au/revealed-true-cost-of-tesla-big-battery-and-its-government-contract-66888/
Additions:
10) The capacity is 900 MWh; delivered, as AC should be added, if known.
– Their maximum charge likely will be at most 60% of that, i.e., 540 MWh. That means 540 MWh/0.96 = 563 MWh enters the battery as DC from the solar plants.
– Their maximum discharge likely will be at most 540 x 0.96 = 518.4 MWh as DC, or 492 MWh as low voltage AC, or 483 MWh as high voltage AC.
11) The battery coupled to solar plants has major losses.
Willem – many good points. It may help to look at the activity of the Big South Australian Battery in context. Here is a chart that shows cumulative net charge/discharge against flows on the Heywood interconnector and the regional reference price (care should be taken to understand that the actual flows on the battery are the rate of change of the cumulative position, which reveals losses in the system as it rises to a multiple of the capacity over the period). At the time, Heywood was thermally limited to 600MW import, 500MW export
This chart shows the output from the Hornsdale wind farm and the charge/discharge of the battery (both as flows, and cumulatively)
It is evident that the battery is not used to stabilise the wind farm output – indeed, you could argue it is destabilising! It is used for response to frequency deviations on the grid – the so-called FCAS market – a service which provides its day to day income.
I would expect that although the battery in Florida isn’t lithium, with its very fast response capability, given that it is just over 2 hours at capacity, it will mainly be used in grid support, with some price arbitrage between peak and off peak. The solar farm is simply a cover story, much as the Hornsdale wind farm is.
It does not add up,
Thank you for the graphs.
The cumulative graph is somewhat useless.
It would be much better if the data showed:
Charge in battery, MWh, at start of day
Charging quantity, MWh, during the day
Discharging quantity, MWh, during the day
Charge in battery, MWh, at end of day
That way, for each day:
Losses can be determined
Level of use of the battery
I think, based on looking at the graph, the battery is an elephant, but does the daily work of a pigmy.
Also it would be interesting the sources and quantities of the revenue of the battery plant.
This facility should be exposed for the expensive hoax it is.
Unfortunately data on the battery charge state is not available to the public: we only see 5 minute data on nominal discharge or charge+own use (for A/C etc.). I found the cumulative line of more use than you, evidently, since it does show roughly what proportion of the storage capacity is being charged and discharged, and it allows some estimate of the round trip losses over time. Because the battery spends a significant portion of its effort on grid frequency support, it can be switching between charge and discharge at frequent intervals, while actually overall quietly charging or discharging: it’s hard to visualise the cumulative position from the original data. I suppose I could show the estimated charge state by only crediting the charge at say 80% of the the power drawn from the grid (roughly the round trip efficiency, which probably varies according to operating conditions – e.g. need for A/C to keep cool). Part of the point was also to illustrate how the battery responds to market price signals. I have another chart on that, which shows some bizarre uneconomic behaviour – discharging at negative prices, sometimes charging at highish prices, etc.
Of course some of this is driven by prearranged testing and by frequency support.
Reneweconomy has some analysis that benefits from some insider information not available to me here:
https://reneweconomy.com.au/tesla-big-battery-it-earned-a-lot-more-money-in-second-quarter-80892/
and here:
https://reneweconomy.com.au/tesla-big-battery-turns-one-celebrates-50-million-in-grid-savings-95920/
even so, it is not gospel.
It doesn’t add up,
Thank you for your additional clarifications and the two URLs, which more clearly describe the HPR operation and the income streams of the HPR.
Most of the revenue is from the FCAS market, which imposes mostly minor charge and discharges on the battery, 24/7/365, with the rest from daily shifting, such as charging during late night/early AM, when prices are low, and discharging during peak hours, when prices are high, aka arbitrage.
Here is an article I wrote, which likely would be of interest to you.
http://www.windtaskforce.org/profiles/blogs/analysis-of-a-6-day-lull-of-wind-and-solar-during-summer-in-new
Good article. I have work in progress on looking at various countries in Europe that I was hoping to offer to Roger Andrews for publication at Euan’s site. Since he died, it has become less active and I’ve not finished the work off or written it up. I should put some effort in: you’ve given me some incentive to resurrect it. I was looking at seasonal and multi year storage requirements, and over-build with curtailment, based on hourly data. I did unearth some work by Dr Friedrich Wagner that suggested that the difference between hourly data and much finer resolution is limited, and likely covered by a sensible margin.
You can access 5 minute Australian grid battery data from this site:
http://nemlog.com.au/nem/unit/HPRG1:HPRL1/20190101/20190409
Follow your nose to learn how to edit the URLs to change the data coverage, and check the nlog and twitter account for many other data options. Easy data download by just clicking on csv.
It doesn’t add up,
Thank you for your additional comments and suggestions.
I have written another article, which likely will be of interest to you.
http://www.windtaskforce.org/profiles/blogs/the-hornsdale-power-reserve-largest-battery-system-in-australia
‘Florida utility to close natural gas plants, build massive solar-powered battery’
They will no longer be powering homes. They will be powering batteries.
I note that griff does his usual spray and walk away commenting. This is often after the big tide of commenters have been through, so his comments appear unchallenged. Either his short comments are unsupported or he links to a favourable press release. All it is really demonstrating is that his searching skills on Google are greater than his actual knowledge on the subject.
Rud: Great piece. However, I’m not sure whether you analyzed the situation from the following point of view: The owner of a battery can buy electricity from the grid when it is in surplus and costs almost nothing. As non-dispatchable renewables become a larger share of generation capacity such situations should arise more often. (Water apparently is being pumped uphill into nearly all dams in the Alps when there is excess renewable power from wind and solar. Arizona is apparently taking some of California’s excess renewable power.) Then the owner of the battery can turn around and sell that power when the price is high.
The competition for the battery owner is presumably a gas peaking plant, which has both capital and fuel charges. For the battery owner, the equivalent to the fuel charge is the cost of the electricity used to charge it – which may be close to zero.
On a mass scale, though?
Would there be that much “close-to-zero” charging electricity to sustain the energy demands of civilization as we know it?
Robert: The law of supply and demand applies to electricity, which is the most perishable commodity in the world. If it can’t be used immediately in the vicinity of its production, it often has zero value because it is too expensive to store or there is limited capacity to transport it over long distances. Electricity from fossil fuel generators never has zero cost, because there is a variable fuel charge associated production and a reasonable amount of flexibility in tuning output to meet demand. This isn’t true for wind, solar or nuclear. When demand is low, electricity from these sources is worth very little. That is why batteries such as these are being installed. That is why it is my understanding that most dams in the Alps has installed equipment to pump water uphill: surplus renewable electricity in Germany is available almost for free.
A favorite anecdote comes from a pioneer “wind farmer” in Alberta. The bearings on many of his turbines were failing and he was asked why he wasn’t replacing them with newer turbines which are more efficient and produce electricity at a lower average cost. He answered: No one wants to buy electricity in Alberta when the wind is blowing. He was waiting for the government to insure a market for any wind power he might produce from new turbines.
None of this makes sense if we are looking down from space and talking about meeting the average (or levelized) cost of production and meeting the average energy demands of civilization. You need to look at the hour=by-hour or minute-by-minute actions of consumers, producers, and regulators that create a market.
Frank you danced around but didn’t really answer the question. “Would there be that much “close-to-zero” charging electricity to sustain the energy demands of civilization as we know it?” Yes they can “charge the batteries” during peak production times (when there’s excess energy) is/will there be enough “excess” energy during those times to cover the times when there is not enough energy being produced to meet societies needs? Perhaps some (made up) numbers will help you understand the question:
Lets say (for simplicities sake) just solar power and we only generate power from the sun for 12 hours a day with a “zero cost excess” energy only being generated for only a couple of hours in the middle of the day (when the sun it at it’s zenith and thus the solar cells are generating maximum energy) – the rest of the time all the solar generated is needed to supply societies needs. Can enough excess during those few peak hours be generated and stored to cover the 12 hours in which we generate ZERO solar energy? Doesn’t seem too likely does it? Which means the battery owners would need to buy a good bit of “non-zero cost” energy to finish charging their batteries in order to cover the non-production times.
John Edicott
It is actually worse than that: On the few days over the entire year when there are “clear skies and low humidity and no dust or pollen”, solar collecters are effective only 6 hours per day! Not twelve, but only the 6 hours between 9:00 am and 3:00 pm. (Now, across most of the world in temperate latitudes, this does average to 7 hours per day summertime, 5 hours per day wintertime.)
So your approximation needs to START with needing 4x the “peak energy” area of solar panels just to provide energy of 1x solar panel average. Assuming no energy lost going into storage, coming out of storage, and then going from that conversion unit through the grid to the load. In practical calculations, you need 6x the area of solar panels needed at noon on the day of your “average” consumption.
It gets worse when you realize that peak demand for electricity does NOT occur at noon (when solar panel supply from the panels is at its daily theoretical maximum) but later, between 5:00 pm and 9:00 pm when almost no solar power is available at all! Between 6:00 am and 9:00 am each mornig when no solar power is available either.