A Freedom of Information Act request to the Michigan Economic Development Corporation has produced 73 pages of records concerning this project. MEDC has published the release in its public FOIA request archive as request 1068, where anyone can download it. The records comprise: the signed Memorandum of Understanding between Los Alamos National Laboratory and the Regents of the University of Michigan, the University’s earlier funding application, an economic impact model, and internal correspondence between U-M and MEDC staff.
These are state records, released by the state. Nothing below is speculation. Page numbers refer to the released document, which you can download from MEDC directly and read yourself. I’d encourage it.
Six things jumped out at me.
1. The agreement describes a facility near Ann Arbor
The MOU was signed by Thomas Mason for Triad National Security, LLC on November 26, 2024 — two weeks before the Michigan Strategic Fund Board approved the $100 million grant.
In it, Los Alamos agrees to support the University’s development of an Innovation Campus in proximity to the University of Michigan’s Ann Arbor campus.
The parties state they intend to work toward establishing a Center for High Performance Computing near UM’s Ann Arbor campus.
(pp. 6–7)
One clarification worth making, because it’s a fair thing to challenge. The Innovation Campus is a separate University project in Ann Arbor, and the MOU treats supporting it as its own area of cooperation. But it also calls the Center for HPC “a central feature of the Innovation Campus,” and says Los Alamos may take office space within it. The computing center isn’t placed anywhere else in that document.
Ypsilanti Township does not appear anywhere in the MOU.
Worth being careful here. An MOU is a framework, not a site plan, and it says outright that it authorizes no work — every actual project needs its own agreement. Plans change for legitimate reasons. But the document these two institutions signed to define the partnership puts it beside the Ann Arbor campus, which raises an obvious question about how it ended up fourteen miles away, on a river, in a township that has formally objected.
2. Michigan was chosen partly for its fresh water
The University’s own application summary explains why the national laboratory selected Michigan. Among the listed factors are the state’s geography and long-term climate resilience, room for expansion, a reliable electrical grid — and ample access to fresh water for equipment cooling operations.
(p. 26)
Access to fresh water for cooling was a reason Michigan won this project.
That’s hard to square with the current public position — that the facility won’t draw from the Huron River or use local groundwater, and that closed-loop cooling, which uses no water, is under consideration. If the design really does end up using little or no water, then one of the stated reasons for picking this state turns out not to have mattered. And if it does use water, we deserve the number now rather than after the design is locked.
The phrase available room for expansion
in the same sentence is worth noting too, given that the site has grown from the 20 acres described to the State to roughly 144 acres today.
3. The 200 jobs, described plainly
We’ve spent a year asking what the “200 permanent research positions” actually are. The MOU answers it in one sentence. Subject to the availability of NNSA funds, Los Alamos anticipates approximately 200 LANL jobs will be created at UM over the next 10 years. (p. 7)
Los Alamos jobs. At the University. Over a decade. Contingent on federal appropriations that have not been made.
Read that against how this has been sold to the township, and against the answer at the September 9 town hall when somebody asked how many permanent jobs would go to residents here — that legally they weren’t sure they could say any would.
4. The state was asked for far more than $100 million
The $100 million grant approved in December 2024 is the end of a much larger negotiation. The records include an earlier ask for a state commitment of up to $500 million, with a recommendation of up to $400 million in state support through bonding and competitiveness funds, conditioned on jobs, investment, federal reinvestment, and a minimum 25% match from U-M. (p. 66)
The application in this release seeks $135 million from the Strategic Site Readiness Program (p. 26). And an indicative state commitment of $400 million is itself listed as one of the reasons Los Alamos selected the University of Michigan.
The same page notes that competing locations were Georgia and Texas.
5. The operating cost estimate is nearly double what the board was told
This version of the project totals $950 million, with a sources-and-uses table that differs from the one presented to the Strategic Fund Board. Construction is $785 million — $685 million for the 100MW building and $100 million for the 10MW building. Equipment for the 100MW facility is listed at $300 million, supported by federal appropriation.
(p. 27)
More striking are the operating costs: $115 million annually for the 100MW facility and $35 million for the 10MW facility — $150 million a year. (p. 27) The December 2024 board packet put ongoing operating costs at $78 million annually.
Those are described as covering debt service, maintenance, general operations, utilities and IT. Electricity is a big share of any number like that. So a community trying to work out this facility’s appetite now has two very different figures from the same institution.
6. The job projections were modeled as a software company
The economic impact analysis in these records projects 540, 790, or 1,040 jobs under low, mid and high scenarios, on an $835 million investment, with jobs ramping up between 2030 and 2034. (p. 62)
The industry used to model those projections is NAICS 541511 — Custom Computer Programming Services.
That’s the classification for a software development firm — a business whose output is people writing code, which is exactly why it throws off high employment and big multipliers in an economic model. A high-performance computing facility isn’t that. It’s infrastructure with a few dozen people in it, and its output is computation. Model one as the other and you overstate both employment and regional impact. Worth asking whether the “500 to 1,000 new jobs” figure going around rests on this assumption.
Two smaller things
Road widening is anticipated. The correspondence discusses a road that will need to be widened up to the main entrance of the facility, with the specifics redacted. (p. 34) No one has publicly quantified that cost or said who pays it.
The site location was redacted. MEDC withheld it on page 26, and the road details on page 34. The location is public knowledge now, so there’s no longer any basis for that redaction. It should be released.
What this does not show
Let me be straight about what these documents don’t prove.
They’re from late 2024. Numbers have moved since, and some of the differences above are probably just a project evolving between drafts rather than anything hidden. The MOU is a framework, not a commitment to build anything anywhere — it says so itself. Nothing in here shows bad faith.
What it does show is that detailed answers to the questions we’ve been asking have existed in writing since 2024, and that they went to the State of Michigan while we were being told this is comparable to a modest industrial building.
What I’m asking for
Four things, every one of which the University could do this week:
- Explain how a project the MOU places near the Ann Arbor campus came to be sited in Ypsilanti Township, and what alternative sites were evaluated — including University-owned land closer to Ann Arbor.
- State the projected consumptive water use in gallons per day, given that access to fresh water for cooling was cited as a reason Michigan was chosen.
- Reconcile the $78 million and $150 million annual operating cost figures, and state the projected annual electricity consumption.
- Confirm whether the publicly cited job figures rely on the Custom Computer Programming Services classification, and if so, publish projections modeled on data center operations instead.
The University held a second town hall on October 1 with the same panel. These are still good things to answer.
Source: Request 1068 in the Michigan Economic Development Corporation public FOIA request archive, comprising the Memorandum of Understanding between Los Alamos National Laboratory and the Regents of the University of Michigan signed November 26, 2024, the University’s Strategic Site Readiness Program application materials, an economic impact analysis, and related correspondence. Comparison figures are from the December 2024 Michigan Strategic Fund board packet; see the sources page. If I’ve characterized any of this unfairly, tell me and I’ll fix it and say that I did.
