[The emailed version of this had typos. Corrections and a few minor editorial changes made here.]
All sorts of schemes and scams pop up everywhere. Some fly, some don’t. Some make sense, some don’t.
Over the years, a long list of schemes and projects have engaged the attention of Green Delaware and other advocacy orgs. Indeed, one of our key functions is to analyze proposals and share our opinions.
How does Delaware decide if a project, or a whole industry, deserves public support? The answer is: we don’t even try.
Numerous agencies and NGO’s are set up to promote whatever scheme is put before them. There are NO agencies or offices charged with evaluating whether a scheme is in the public interest. Let me say that again: There are NO agencies or offices charged with evaluating whether a scheme is in the public interest.
Occasionally an elected official or agency will try to exercise responsible judgement; ask the hard questions. Rep. John Kowalko (Newark) does this consistently, and takes a lot of flak for it. When John Hughes headed up the DNREC, he was supportive of opposition to the DuPont/US Army scheme to dump poison gas residue into the Delaware River. It made a big difference and the scheme died.
A few years ago NRG proposed a new coal burner at the Indian River Power Plant in Millsboro. Most politicians and unions signed off on the idea immediately. Public opposition was broad and deep. Thankfully, this project died.
Around the same time, an offshore wind project near the mouth of the Delaware Bay, Bluewater Wind, was proposed. Public support was broad and deep. Green Delaware supported the project quite vocally but stood alone in also saying a rigorous permitting process was needed because of many unknowns. The Bluewater Wind people responded by giving money to stooge “enviro” groups. Rigorous review never happened. The project has never been built.
Some years ago people were promoting a coal-fired corn ethanol distillery near Claymont–illegally in the (supposedly) protected Coastal Zone. This quickly died of foolishness.
Green Delaware is following several schemes on the loose in Delaware these days, including a tire incinerator near the City of New Castle, two chicken litter incinerators in Sussex County, and “The Data Center” in Newark.
How to decide if a scheme is a good idea?
Well, if you are Governor Markell or the Delaware Economic Development Office (DEDO), or the Delaware Department of Natural Resources and Environmental Control (DNREC), or the Delaware State Chamber of Commerce, or the Editorial Board of the News Journal, or the Delaware AFL-CIO, the question is simple: EVERY scheme is a good one, given only that it promises “jobs, jobs, jobs” (very publicly), “growth” (presumed always to be a good thing) and profits for property owners, investors, and various parts of the financial services industries (usually not mentioned so publicly). Who cares about the environment or the quality of human life in Delaware…?
But if we are concerned about the impacts and costs to society, about limiting climate change so human life can continue somewhat as accustomed, and so some of Delaware can remain above water, about public health, quality of life, residential property values, and so on, we need to make more thoughtful decisions.
What criteria should we use to decide if a project, or a whole new industry, is desirable? What role should the public have in the decision? Is Delaware’s political system set up to consider the public interest? To give the public (us) a say?
The so-called “Data Center” scheme in Newark is interesting because it’s apparently not a “data center” at all. It’s mainly a power plant. It’s interesting because all sort of promises are being made that seem unlikely to be kept. And it’s interesting because the numbers keep shifting. And because state money has been promised to the project. It’s interesting because we are hearing reports of threats and intimidation of small businesses and others, and because The News Journal is allegedly refusing to print most letters from opponents.
Data centers typically do need backup generators, and this has been cleverly used by the promoters to muddy the waters of public and official perception.
Lots of interests–unions, various Republican operatives, the City of Newark, the University of Delaware, the Dealer in Chief–are banging the drum for the “Data Center” but as far as I can see none know much about it. “Due diligence” does not seem to apply.
The Data Centers, LLC, is apparently located in West Chester, PA. The operation seems not to have any listed phone numbers. This contact number is found in the air permit: 610-675-6230
What’s a “data center” (sometimes called a “server farm”)? Basically, climate-controlled buildings filled with computers on racks. Since data centers are “mission critical” for their owners, and even a brief power interruption can crash the computers, reliable, backed-up electric power is a key need.
Grid power is highly reliable in the US, with average “uptime” being in excess of 99 percent. (There are 8760 hours in a year, so if the power is off 2 hours/year the uptime of the power supply is 8758/8760=99.98 percent). But outages do happen and data centers can’t tolerate that. They need backup power, which usually takes the form of battery backup systems able to carry the load for a few minutes or hours, and standby generators, usually Diesel, able to carry the load for perhaps a few days. The backup systems are similar, except in scale, to the “UPS” system you might have sitting under your office desk. The battery system can carry the load until the generators start. But some data centers, apparently, run generators a lot, or continually, and many have been cited for air emissions violations. Some data centers use spinning flywheel backup systems.
How much power a data center needs obviously depends on the size, and is sometimes called out in watts per square foot. In general, it appears that the amount of electricity used by the computers is about equal to the amount used for air conditioning–keeping the equipment cool.
How much power is actually being used by data centers is unclear but the New York Times reported a year ago that the proportion used for actual computation averaged 6 to 12 percent. This suggests the possibility of massive improvements in energy efficiency.
A typical “large” data center in 2008 was described as 50,000 square feet and 5 megawatts (MW). The same source forecast that a “large” data center in 2020 would be 500,000 square feet and 50 MW. The largest data centers I could find reports on had power supply ratings in the vicinity of 100 MW. For context, the total peak load of the Newark Electric Department is about 92 MW.
However, a 2002 report by Lawrence Berkeley Lab found the actual electrical consumption of a number of data centers studied to be less than 40 watts per square foot. The actual consumption was less than the ratings of the supply systems.
The proposed Newark “Data Center,” would, according to a document on the promoters’ website, be 288,000 square feet. Using the reported 1988 energy consumption of 100 watts/per square foot, that gives us a load of 29 MW. Using 40 watts/square foot gives 12 MW. I’ll be conservative and use the 100 watt/square foot figure.
The Data Center says, in the same document:
a. 108 MW is required to support the computer hardware located at the facility
b. 90 MW is required for data center cooling and ancillary equipment, such as lighting
c. 50 MW is required for backup in the event a turbine or engine fails
Total: 248 MW. 8.5 times a reasonable estimate of the actual load. Disregarding the “backup” capacity, this seems to equate to an energy density of 687 watts per square foot.
But, the actual air permit application submitted to the Delaware Department of Natural Resources and Environmental Control says, at page 2:
” … aggregated combustion units (gas turbines and reciprocating engines) and steam turbines producing electricity will have a combined maximum nominal power capacity of approximately 279 MW, accounting for the capacity of the gas combustion turbine held in reserve. The planned operating electricity output of the CHP, based on six combustion turbines, three reciprocating gas engines and the three steam turbines will be approximately 181 MW. Approximately 131 MW of electricity is required to operate Wolf 1. Approximately 50 MW is anticipated to be supplied to the City of Newark (through DEMEC) under normal conditions.”
279 MW is almost ten times the reasonable load.
131 MW is different than 198 MW
What’s this all mean? Which of these various conflicting numbers, if any, should be believed?
Note that the grid is far more reliable than any one generating unit, or even several clustered together as proposed here. the “PJM interconnection,” the “grid” in this part of the country, is fed by hundreds of generating units, no one of which is vital. Transmission lines as such, with all their disadvantages, are far more reliable than generating units.
“The Data Center” folks say they are going to forgo all the advantages of the grid and operate as an “island.” “TDC data center will not draw its power from a local electrical grid (a network for supplying energy), rather it will supply and meet its own energy requirements, standing alone as an ‘island.'” Yet “The Data Center” apparently also plans to supply the City of Newark, and this raises questions about interconnection with PJM, a complicated procedure.
Are “The Data Centers” really going to build 279 MW of capacity to support a 29 MW load? How would this be in the public interest? How could it be profitable for investors? Whatever is really going on here seems to be different than what the promoters are saying.
More information upcoming in a couple of days on the political and regulatory side of all this. Why is the DNREC seeking to conceal information? Why is community opposition so strong? Why is the City of Newark making the absurd claim that the power plant is an “auxiliary use” to the Data Center, and would not require approval from the City Council? Lots of questions to explore….
Thanks for reading.