Building on Virginia’s Data-Center Boom

Data centers are the hottest trend in Virginia economic development these days. But the state is only beginning to think through the implications.

Loudoun County, home to 75 facilities, has developed the largest cluster of data centers in the country (and perhaps the world), and next-door-neighbor Prince William County is rising fast. Rural Mecklenburg County has attracted nearly $2 billion in investment as the location of Microsoft’s East Coast hub for online services. QTS has retrofitted an old microchip factory in Henrico County to open a data center, while DP Facilities, Inc., opened a $65 project center in Wise County. Soon, Virginia Beach will enter the data-center sweepstakes when construction is complete on a 4,000-mile transatlantic cable connecting Virginia to Europe.

According to Paula Squires writing in Virginia Business magazine, Virginia boasts more than 650 data processing, hosting and related establishments that employ more than 13,900 people. Since 2006, the industry has announced more than $11.8 million in new investment and 6,600 jobs. The jobs, while relatively few in number, pay well (more than $100,000 a year in Northern Virginia), and generate a gusher of local taxes.

Billions of dollars are flowing into the sector as the global economy embraces cloud computing to handle the massive surge in data collection and storage. A Markets and Markets research report estimates that the cloud storage market will grow from $23.76 billion in 2016 to $74.94 billion by 2021 — a compounded annual growth rate of 25.8%.

Loudoun County was one of the first localities anywhere to see the economic development potential. The county had a built-in advantage — a massive network of fiber-optic cable built by AOL and WorldCom during the heyday of the 1990s Internet bubble. WorldCom went bust and AOL has a much-diminished presence, but the cable infrastructure remained — and high-capacity connectivity is an essential prerequisite for a data center. Loudoun claims that 70% of the world’s Internet traffic passes through the county. The concentration of data centers is so pronounced that economic developers refers to a six-mile radius around Waxpool Road and Loudoun County Parkway as “data center alley.”

The county has built on its infrastructure advantage by learning how to expedite zoning, permitting and construction. CyrusOne completed construction of a 220,000-square-foot data center in Sterling in 180 days — reputedly the shortest construction time fever for a center that size, reports Squires.

To incentivize investment, the state exempts computer equipment bought or leased for a data center from the retail sales and use tax. Henrico County has dropped its business property tax rate on computers and related equipment from $3.50 to $.40 per $100 of assessed value.

Also, Dominion Energy has emerged as a significant partner. The endless racks of servers inside data centers consume electricity and generate heat, which must be cooled by massive HVAC systems. Dominion charges 5.2 cents per kilowatt hour for large facilities, and a slightly higher rate for small ones. “We’re very competitive,” says Stan Blackwell, director of customer service and strategic partnerships for Dominion. “We have some of the lowest data-center rates in the nation.”

Bacon’s bottom line: The rise of the data-center industry raises two pointed sets of public policy questions.

First, how can Virginia optimize this opportunity? What are the critical drivers? Obviously, the existence of high-capacity fiber networks is one consideration. It appears from the map atop this post that Virginia has one of the densest clusters of long-haul fiber capacity in the country. How crucial is that advantage? Does Virginia’s proximity to a relatively fiber-poor Southeastern U.S. give data centers serving that market an edge? Is the competitive advantage bequeathed by fiber-optic infrastructure such that Virginia should consider encouraging investment in more? Conversely, does it do any good for Virginia to invest in its own fiber infrastructure if connections to neighboring states are lacking? Many, many questions.

Electricity is one of the largest costs associated with operating a data center, accounting for roughly 10% of the total cost of ownership — and it is one of the largest costs that vary by location. Dominion’s electric rates confer a significant competitive edge for locations within its service territory.

Graph credit: Dominion Energy

One of the biggest challenges for Dominion — and the further expansion of the data-center industry — is delivering electricity to these data centers. In one particularly controversial case, the utility wants to build a 230 kV transmission line and substation from Gainesville to Haymarket to serve an Amazon data center. Locals have organized in opposition, claiming that the 100-foot-tall towers will disrupt views and harm property values to benefit a single industrial customer. They insist that Dominion bury the line at considerable expense. If Virginia wants to develop the data-center industry more fully, it may need to find ways to resolve the inevitable utility-landowner disputes fairly expeditiously. No company wants to wait years to find out whether a project will get the electric power it needs.

A second big public policy question centers on the implications of the data-center boom for electricity demand in Virginia. According to Virginia Business, data centers represent Dominion’s fastest-growing customer segment: About 7% of the company’s retail portfolio consists of data centers.

This feeds into the debate over Dominion’s future electric generating mix. Dominion’s 2017 Integrated Resource Plan (IRP) assumes that electric load will increase at a compounded rate of 1.5% over the next 15 years — considerably higher than PJM Interconnection’s forecast for the Dominion service territory. Dominion argues that PJM has not taken into account the phenomenal growth of demand by Virginia-based data centers. These projections matter because they influence how much new generating capacity — including nuclear, as I will explore in a forthcoming post — Dominion adds in the years ahead, with tremendous implications for rate payer and the environment.

The data center surge could prove to be an economic development boon for Virginia. But the industry’s growth impacts local zoning and land-use practices, tax policy, fiber-optic infrastructure development, and energy policy. The McAuliffe administration would be well advised to pull together a conclave to determine how to sort through these issues.

There are currently no comments highlighted.

25 responses to “Building on Virginia’s Data-Center Boom

  1. Don the Ripper understands the Cloud better than anyone else who frequents this blog. I hope he will share his knowledge about the factors driving locational decisions for new data centers and how Virginia stacks up against other states.

  2. …another question would be – are the cloud participants demanding Virginia increase cost for electric power – by going to renewables only? In other words are they saying? “the Data Center industry does not want to build or pay for our own renewable/solar electric facilities, but we demand that Virginia do so, or else we leave the State”…because that’s the ultimatum many Democrats insist they are hearing.

    My guess the energy analogy is closer to the airline industry, and they want their jet fuel to be as cheap as possible. In which case Virginia should be in pretty good shape.

  3. Either you don’t understand the Haymarket Power Line issue or you are simply providing cover for Dominion. The “Locals” aren’t claiming it will “benefit a single industrial customer” that is a fact. When Dominion filed the application it asserted that 75% of the capacity would be used by Amazon. When the public hearings were complete, they had raised that to 97%, the numbers (their numbers) tell the story.

    As to the claim that “100-foot-tall towers will disrupt views and harm property values”, when those towers are proposed to be erected through subdivisions, on top of wetlands and through battlefields, you’re damn straight skippy.

    As to the cost, as with virtually every other comparable circumstance, Dominion has grossly overstated the cost and maintenance expense of burying the lines by relying on outdated cost estimates based on outdated technology and practices. At the same time, Dominion grossly understated the above ground cost by conveniently forgetting to include most of the condemnation costs, mitigation costs, etc. To be frank, their estimates are greater works of fiction than those available on Amazon.

    The difference in this particular instance, the difference that has to be driving Dominion bat-shit crazy is that the “Locals” are comprised of a highly educated and professional demographic with considerably more experience dealing with the Gubmint than practically any other jurisdiction in the county. That’s what happens when you site those facilities in a area where you can’t swing a dead cat without hitting a GS-15, SES, General, Law Partner, CEO, engineer, MBA, Research Chemist, etc., etc., etc. up to and including former public utility engineers and executives.

    You are correct in one thing, Virginia does need to find a way to resolve the inevitable utility-landowner disputes expeditiously, those “Locals” have put Dominion at least two years behind schedule on this project and I suspect that if they do ever turn a spade of dirt, it won’t be until well into the next decade and likely not in a location of their choosing.

  4. So why can’t Mr. Big Bucks Bezos build sufficient solar panel capacity on his data center campus to power all or a major part of his data center. It’s like Mike Bloomberg donating a paltry $15 million to a climate change counter-initiative. Many of the wealthy talkers aren’t into the walk of their talk.

    Back during the Civil War, many wealthy people would singularly or with friends raise and fully equip as much as a regiment – 1000 men plus officers in the day – from their own funds because they believed in the Union or the Confederacy.

    • “So why can’t Mr. Big Bucks Bezos build sufficient solar panel capacity on his data center campus to power all or a major part of his data center.”

      Now you get to the truly fun or diabolic part of the matter, depending on your viewpoint. When Dominion filed the application in 21015, the necessity was described thusly: “The electric facilities proposed in this application are necessary so that Dominion Virginia Power can provide service requested by a retail electric K service customer (the “Customer”) for a new data center campus in Prince William County” (Paragraph 2).

      That was then elaborated on in Appendix I: “The total Customer load at Haymarket Campus is projected to be approximately 120 MVA, consisting of three buildings. The proposed new electric transmission facilities must be in service by June of 2018 to serve the Customer’s new development. The total loading at Haymarket Substation, including the Customer’s load, is projected to be approximately 160 MVA at full build-out”.

      Fast forward to March of this year and at a COE Section 106 meeting regarding permits, we have Amazon/VAData’s attorney’s from Williams Mullen stating for the record that:
      1. Building One is complete and operational using existing electric infrastructure.
      2. Building Two could operate without additional electrical infrastructure.
      4. The Haymarket Power Line would not be required until Building Three was operational.
      5. Buildings Two and Three were not expected to be built in the near future if at all as their construction would only occur if expanded capacity were required.

      So the necessity to justify the power line by the in-service date of June 1, 2018 (that pesky three bulding data center) had shrunk to one building operating on existing power by March 8, 2017. Funny how that assertion is made nine months after the public hearing and four months after the hearing examiner’s report was issued, too late to be entered for consideration by the SCC.

      Funny thing, although those assertions came as quite the surprise to the consulting parties present at the meeting, they would appear to have come as an even greater surprise to Dominion as one could actually hear the sphincters of Dominion’s attorneys pucker as the attorneys from Williams Mullen spoke.

  5. This is an intriguing issue. The new data center loads comprise most of Dominion’s load growth, with a small amount contributed by expected population increases. The difficulty is that data centers want to be be provided with electricity generated by renewables, not gas-fired plants. Other than the fact that solar facilities are now cheaper than gas-fired plants, I suspect this is a primary reason Dominion has shifted so heavily from gas to solar in this year’s IRP.

    Data centers are not a huge contributor to economic development. They do not require a large workforce to operate. Other than a short-term blip to construct or renovate a building I wouldn’t think that they would provide a noticeable increase in economic activity in NOVA even with 75 of them.

    Although the “cloud” is the latest rage, there are some considerations affecting its future expansion. The Internet of Things (IoT), the next big thing, is encountering difficulty with the “cloud”. When cloud service is temporarily unavailable, people are finding that with the new home energy systems, they can’t change their heating and cooling settings or turn a light on, because current designs all run through the cloud, not with local storage and control. How this might affect development of more data centers is not clear.

    Amazon and others have met their desire to be served by renewables by buying Renewable Energy Credits (RECs) rather than relying solely on actual renewable generation. Although some, like Apple, are providing more of their own renewable generation.

    Some onsite solar generation on building rooftops and parking lots could meet a portion of the load, but probably not enough to serve all of the load. This is more the domain of third-parties and not Dominion and the utility doesn’t want to lose the revenue. I’m not sure how well designed these facilities are. They have lots of money and very bright people. But there is currently a huge heat load from all of the microprocessors. Innovative layouts to more efficiently disperse the heat have been proposed, but I don’t know if they are widely adopted.

    We have come close to the lithography limits of current chips designs (although layers are now being stacked to reduce electron friction and heat). We might have to wait for photonic computers or some other low-heat design to drastically reduce the electricity requirements of these facilities.

    Data centers and utilities are two huge command and control, central station type operations. It will be interesting to see how the disruption due to distributed energy will affect their business models. The need to concentrate around the big fiber network might over-ride any other decisions for now.

    • While I don’t think the Cloud will go away, it’s just a swing in the pendulum from centralized network control to putting intelligence at the edges of the network. Over the last 40 years, I’ve seen lots of swings.

      I remember time sharing on main frames (college projects), centralized word processing through Wang terminals, and host-remote telco central offices. While the intelligence remained in the switches, Advanced Intelligent Network was an attempt to put service creation control in the hands of the end user or company network administrator. And I’ve used many personal computers loaded with software as well as the law firm’s current use of cloud-based software.

      If you centralize intelligence or storage in single location, you have a single point of failure, such that when its out of service, so is everyone else. That’s why telcos went to self-healing fiber rings and duplicate SS7 facilities years ago, and why most IP-based carriers use route and facility redundancy. As I understand network theory, signaling messages, packets and calls all move from the failed point to find working paths, which, in turn, can overwhelm the newly found paths. This cascading of 1s & 0s can shut down the network, as found in a few major SS7 failures in the 1990s.

      A sound next step in my mind is centralization of control and operations in the Cloud connected by 5G wireless facilities, but with the ability for users to shut off centralized services and take manual control. This is not unlike our large jet airplanes where pilots can turnoff autopilot.

  6. hmmm….doesn’t sound like Loudoun wants those data centers to start with, eh?

    You can bet any companies waiting in the wings are going to look elsewhere…

    I do suspect that as fast as Loudoun has grown – they probably need more power now also..
    ironic the seeming connection between govt employees and NIMBY!!!

    why… I bet all those Jamestown waterfront NIMBY folks are RETIRED GS-15s!

    so what is the essential issue.? that no new powerlines are needed or that they are needed but need to be done different than proposed?

    we’ve have issues like that down our way too and back and forth about need and doing it “different” but in the end – it basically was the folks who would be near the proposed lines that were opposed… and mostly on aesthetic grounds.. and Dominion did not want to bury them.

    so.. we can’t have powerlines, or solar farms, or pipelines or wind turbines or cell towers , etc, because they crap of the view and they all need to be underground?

  7. Pingback: Dominion Energy (D) Receiving Favorable Media Coverage, AlphaOne Reports - Stock Observer

  8. data centers are just infrastructure… like an electrical substation or a solar farm, etc…the actual use of them in doing “work” is remote..to the site.

    lots of servers and storage with redundant fail-over units running 24/7 needing electricity and a few IT professionals to maintain them unless something breaks. they’re not labor-intensive job creators.

    Was amused to read that the recent British Airways outage was caused by someone turning off the power…hard to conceive!!!

    TomH alluded to the IOT issue with remote servers or internet going down and the thermostat I have runs with local control if it loses internet connectivity and I expect most IOT stuff to be designed to run this way by default.

    but the whole IOT thing also relates back to future demand for electricity because IOT is among other things – demand side control of electricity use and as it spreads, I’d expect it to have a profound affect on reducing electricity demand.

    interesting article in local FLS paper on solar panels – and another in WSJ on Tesla Powerwalls now being installed in tandem with solar roofs.

    ” Home Batteries Aren’t Economical—Yet
    Homeowners can now power their homes with batteries, but the benefits are minimal in most areas for now”

    https://www.wsj.com/articles/home-batteries-arent-economicalyet-1495418520?mod=e2twd

    the roof panels pay for themselves in 15 years or so.. with people reporting their electric bills dropping to 1/3 or less even without selling the excess back to the utility. If new houses are built with the solar already and the price embedded in the mortgage – combined with powerwalls and IOT , in the next 10 -20 years – I don’t know where Dominion is getting their “projections” for more electricity need… but I do acknowledge that they know their own business and would not be building more plants without good reason.

    Finally, I note they did get approval of NA3.. and according to what I read are “good to go” from the Federal regulators… but still need the SCC to weigh in.

    I have to say – as bad as nukes are in cost and waste disposal – that if the alternative to not having enough baseload is to burn gas as baseload with it’s potent greenhouse gases, I’m seriously conflicted.

    perhaps solar and powerwalls will come on quick and strong and save us…

  9. There is a lot of discussion about lack of reliability as a result of higher penetrations of renewables that is unfounded. Every ISO in the US has a significant surplus of generation. PJM has a 28.8% surplus and needs only about half of that for reserve capacity. More combined cycle units are on the way, far more than are justified by increases in load growth.

    We not in any near- or medium- term danger of not having enough capacity.

    Larry is right, numerous innovations are on their way that will improve efficiency and demand side management. In Virginia, if we just improved energy efficiency by 1.5% of our load each year, we would not need to build any new baseload facilities at least until 2032 and still maintain the same or greater reliability as we have today. Several states are currently improving efficiency by over 3% per year right now.

    Distributed solar panels and solar roofs will also diminish the need for new central station generation.

    The rush to build new power plants and pipelines is unnecessary, expensive and environmentally destructive. We are foreclosing a cleaner, cheaper future by extending the 20th century far into the 21st.

    Tucson Electric just signed a contract for solar plus storage for 4.5 cents per kWh for 2018 delivery. We are not in Arizona, but this price point will just be a few years down the road for us too.

    The essential requirement for a modern energy system will be flexible generation that can adapt to changes in load and variations from renewable output. Nuclear and coal units do not meet those operational requirements, or produce energy at a price that will clear the market auctions. Combined cycle units can meet those requirements to some degree, but will generate more expensive energy as gas prices rise. We will have a system with a blend of nuclear, coal and natural gas units for several decades, but we should be very judicious about how many and what type of new units we choose to build.

  10. Pingback: Dominion Energy (D) Receives News Sentiment Rating of 0.43 - BBNS

  11. In a perverse way – actions taken that will increase the price of gas ergo electricity will more quickly spur solar and demand-side efficiencies!

    A cynic might observe that Dominion is doing what it can to sell more electricity at higher prices – and make it harder and more expensive for consumers to adopt solar and demand-side efficiencies.. but the cooperatives seem to be on the consumers side..

    • The co-op on Kaua’i is like that too. But it helps that the owners are your customers. It is easier to get an alignment of interest with a co-op than with an investor owned utility (IOU). But comparing how the Kauain co-op, KIUC, is embracing solar and storage compared to the other investor-owned utilities in Hawaii, that are putting limits on how much and in what way solar can be connected in their systems, illustrates the dampening effect that shareholder concerns have on innovation.

      That’s why I keep preaching that we need to change the rules in Virginia so that by serving the needs of customers better, utility shareholders will also be rewarded. Currently, the decisions are heavily slanted in the shareholders favor. Even if the SCC fully recognizes that (they haven’t), it would be difficult to strike a proper balance with purely a cost-of-service regulatory scheme in a 21st century energy system.

  12. This says it all: (copied from earlier in this discussion)

    5.2 cents versus 8.6 cents is a HUGE advantage for Dominion when it comes to attracting data centers, which are HUGE consumers of raw electric power 24 hours a day.

    Yet you say, “A cynic might observe that Dominion is doing what it can to sell more electricity at higher prices.” I don’t get your logic. Dominion is a low cost provider. Dominion is doing everything it can to stay that way! Including, taking advantage of the cheap gas prices today to keep its prices low, along with the risk that gas prices will rise sometime in the future before the rash of new gas units are fully paid for.

  13. Acbar,
    I was going to ask Jim if this was really the data center rate or Dominion’s basic industrial rate. I think Dominion was going to create a special tariff for the data centers that was based on actual renewable generation not just Renewable Energy Credits (RECs). It is not clear from this article what the data center rate is.

    Amazon, Google, and others have been clear that they do not want to be supplied by gas-fired generation, although in other states they have used RECs to have the appearance of clean energy. I thought they entered into a PPA for the new solar facility that Dominion bought.

    I do not know if the 5.2 cents includes all of the Rate Adjustment Clauses and wires charges, as well. The 5.2 cents could also be a below cost rate subsidized by residential customers as it is in some states. It could also be a rate that is in addition to the demand charge that makes the energy appear cheaper than it actually is. Every state handles their commercial and industrial rates differently, so it is difficult to make a direct comparison.

    For example, the RGGI states are all in the Northeast with much higher rates in general, which also pulls up the East Coast average. I think it is the concentration of fiber more than an electricity rate advantage that draws them here.

    • Good point about the rates; I don’t know what’s in the 5.2 cents or what Dominion’s special data center rate is, if any. DOM’s energy generation over a given time is whatever PJM chose to dispatch, and DOM’s annual and short term capacity sales are what they are.

      When Amazon says “they do not want to be supplied by gas-fired generation” but want to be connected to the grid, you and I both know that the electrons that pass through those data center wires are commingled electricity from all grid sources. Yes, it’s true that Amazon has entered into solar generation deals (e.g. the Accomack installation Jim has talked about) which probably (I have not seen the contracts personally) calls for solar generation into the PJM grid when generating with energy sales by that generator during those times directly to the PJM energy market. Those sales would also create RECs which could be sold directly to the retail data center customer. It’s important to understand, in PJM (and I believe throughout the U.S.), one REC is created per megawatthour of renewables electric energy generated — which is to say, you can consume RECs at night that were created entirely by solar generators. A data center which operates 24-7 can be supplied sufficient RECs entirely from a solar generator which only generates during daylight, as long as you create enough during the day to “bank” them for night consumption. RECs are not time-specific; obviously electricity does not come 24-7 from a solar generator that only works in sunlight. If a retail data-center customer truly wanted to get its electric energy on a simultaneous, generated-over-here exactly equals consumed-over-there, basis it would have to enter into a requirements wheeling contract with PJM to deliver from a specific generating source (PJM does offer a point-to-point wheeling service as required by FERC, but just about nobody uses it, and if they did it would be for a steady quantity over time; I never even heard of a variable or “requirements” wheeling arrangement). The impracticality of this is why REC markets were invented.

      Now it’s possible that DOM has got involved as a middle-man and is buying sufficient RECs from PJM via its GATs Bulletin Boards or from brokers who handle RECs that they can advertise a special rate schedule for a supply of retail electricity that’s 100% “from” renewables. I’m OK with that, as long as everyone understands that a REC is not time-of-day specific and may have come from any type of renewables generation at any time in the past.

      • re: Is Amazon really buying non-fossil-fuel electricity.

        I won’t presume to know Amazon’s thinking but one way of
        thinking about it (besides self-aggrandizing PR) , is that buy stipulating in contract – they are essentially insuring that someone, somewhere, will put up the about of solar they are signed up to get.

        If not mistaken – isn’t that exactly what Amazon did initially on the Eastern Shore?

        https://communityenergysolar.com/project/amazon-solar/

        and they buy ENOUGH of it to power 24/7 which means enough additional during the day – over and above what they need during the day, to be enough “watts” that they would need at night.

        In other words, they buy more than enough during the day – to offset fossil fuels during the day – so they can then “use” the offset fossil fuels at night.

        Make no mistake – at some point if “storage” becomes “real”.. they’ll do that… instead… but perhaps not on-site. Amazon does not want , itself, to get into the business of providing electricity per se…on site for themselves – yet…

        unlike WalMart:

    • A little more research turned up this chart:
      https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0ahUKEwi42vy83KLUAhWE5SYKHaoPA_wQFggpMAE&url=https%3A%2F%2Fwww.dom.com%2Flibrary%2Fdomcom%2Fpdfs%2Fvirginia-power%2Frates%2Fvirginia-rate-updates-and-charts%2Findustrial-nat-avg.pdf%3Fla%3Den&usg=AFQjCNEErybs-JPzs9TF-biN7oZmFcsxNA&sig2=RPk04kN5XOw5G02XmUjAsQ
      This is not a .jpg so it won’t reproduce here but you can view it yourself online. This chart indicates that the 5.2 cent rate is the current “typical industrial” rate for a Dominion retail customer having “1,000 kW demand and 650,000 kWh usage, monthly, annualized.” Thus the rate includes all riders and demand charges and is a weighted average of all different times of day etc. The comparison regional and national rates are calculated the same way and all are compiled in an annual report published by EEI.

  14. re: ” Dominion is doing everything it can to stay that way! Including, taking advantage of the cheap gas prices today to keep its prices low, along with the risk that gas prices will rise sometime in the future before the rash of new gas units are fully paid for.”

    I was not clear…

    but here’s how you increase costs:

    1. build a 5 billion dollar pipeline for gas when there is existing gas for far less cost.

    2. burn gas for base-load rather than peak load which will draw down the reserves quicker.

    3. build a nuke for 19 billion and pass that cost onto rate payers.

    so if solar comes on strong and people start using to lower their electricity costs and thus demand – will Dom go to the GA and say they need to recover their stranded investments?

    So is Dom “planning” to keep prices low ?

    • They might well try to recover their stranded costs, but they would have a poor case to make for doing so based on your assumptions. As for building the nuke, the assumption (as Jim clearly articulates here) is that they feel they have no other reasonable choice and that the SCC agrees. No. 2 is a baseless distinction today; natural gas comes out of the ground with no such restrictions on use and the last attempt to mandate such a difference (under Jimmy Carter) was repealed in the 1990s (under Clinton).

  15. well the point is they’re going to build an expensive pipeline to sell gas to themselves and the cost of it will likely be more than from competitors who already have a pipeline well amortized …

    so that’s not going to contribute to lower cost electricity..

    and if supplies of frack gas start to dwindle ..the gas at the wellhead is going to cost me… and snowball as it gets transported to the gas plants.

    finally NA3 is going to add a lot to the cost of electricity if built.

    bottom line – I don’t see what Dominion is doing to insure that the cost stays low… I just see things that likely are going to increase costs…

    their cheapest path is to use solar when available and gas when it’s not.

    burning gas 24/7 as base load while discouraging solar… is not going to lead to lower costs.

    right?

    • We agree on this; don’t read too much into what I’m about to say. But Jim is correct, IF the State adopted the mass-based CPP option, IF the CPP survived the political onslaught, and IF everyone turned to gas and the price shot up, and IF nuclear fuel and storage costs miraculously came down, then NA3 might become an option worth having kept alive. And once cancelled, it probably couldn’t be revived. But I think the truth also contains the financial aspects Steve Haner mentions above. And we both know NA3 is dead as a doornail anyway as a practical matter; so why continue to bother to evaluate it as though it were a serious option, other than to save face (and costs already sunk)?

    • One other comment: you say, “burning gas 24/7 as base load while discouraging solar… is not going to lead to lower costs.” Not that simple. Base load is the 24/7 load, NOT the amount above that which ISN’T there 24/7. As TomH says below, the bigger risk is not that solar won’t find a market, but that it will continue to take on greater market share and displace so much of the daytime market for gas that gas units built on the assumption they’d serve load 24/7 will be cycled off most days, which is less efficient for them and may even mean they won’t run enough to survive financially. I’m not losing any sleep over that possibility, but already most gas units being built today are not being built as combined-cycle plants but built to cycle daily, at a cost in maximum efficiency, even though in the short run they probably will run 24/7. As TomH points out, this problem won’t go way until the ability to time-shift solar power through batteries becomes cheaper.

  16. The cost of energy has gone up every time a utility builds a new conventional power plant since the mid-1970s. This is not just true of Dominion, but all utilities. As in Virginia, base rates might not change, but RACs are approved every time a new conventional power plant is approved.

    The decline in energy prices over the past few years was due to lower natural gas prices, something the utilities had nothing to do with. Utilities did add a higher percentage of gas-fired generation and low-cost renewables were also used to a greater degree.

    Natural gas prices have increased nearly 100% in the last year. The forces that led to the decline in gas prices are unlikely to continue, except for technological improvements in drilling which contributed about 20% of the cost savings.

    Well productivity is declining, so gas prices will continue to increase, especially as we begin to send much of our cheap gas overseas. The payback of fixed price options such as solar and energy efficiency will only get better as the cost they are competing against rises every year.

    By the way Larry, using gas-fired generation to meet the base load does not discourage solar. Solar will keep getting cheaper and will be used more often. The risk is to the gas-fired unit if solar penetration gets high enough to displace gas from meeting some of the baseload and the gas-fired unit is dispatched less often.

    A new combined cycle plant cannot generate energy at the the 4.5 cents per kWh of that new solar plus storage plant in Tucson, but current lithium ion batteries cannot store and release energy for long enough to fully meet the nighttime load.

Leave a Reply