MISO's Fast Lane for Data Centers That Bring Their Own Power Draws Support From the Data Center Coalition and a Protest From Google

The trade group for data center owners told FERC on Friday that MISO’s new study for big loads with their own generation should be approved. Google, filing the same day, told FERC it can’t be approved yet. The proposal is the Midcontinent Independent System Operator’s Large Load Addition Resource Study, or LARS, in Docket ER27-25-000, and comments were due at 5 p.m. Eastern on October 23 ^[2]. The load at stake is a floor, not a project: LARS is open to any single load of at least 200 MW ^[1].

The split is not over the idea. Both the Data Center Coalition and Google say studying a load and its paired power plant together is the right approach ^[3,4]. They differ on whether FERC can approve the study before MISO files the rules for who pays.

What MISO filed

MISO filed LARS on October 2 under Section 205 of the Federal Power Act, asking for a December 2 effective date so the first study period can open in the first quarter of 2027 ^[1]. The process is permanent. MISO describes its aim as to “facilitate the cost effective, timely and reliable integration of new large loads into the system” ^[1].

The mechanics, from the transmittal letter ^[1]:

• Who applies: the “LARS Associated Transmission Customer,” usually the load-serving utility. Only it can submit. The generator, the transmission owner and the load’s owner fill in their parts of one joint application.
• Size and location: one load of at least 200 MW at a single withdrawal point, paired with new generation of at least 200 MW in aggregate, in the same Local Resource Zone. Generation is capped at 150% of the load’s peak demand and limited to network resource interconnection service.
• Contracts: a binding agreement of at least five years at application, a commitment to stay paired for 15 years, and a commercial operation date within three years.
• Money: a $250,000 non-refundable deposit and a $30,000-per-megawatt entry milestone, refundable in full on withdrawal before the interconnection agreement is signed.
• Volume: ten applications per study period, three periods a year, each finished in about 120 days.

MISO first proposed a 250 MW floor. It cut it to 200 MW after stakeholders said data center and industrial projects in the 200-250 MW range also have significant system impacts ^[1]. Its own expedited-review results, MISO wrote, show upgrades “increasingly likely” between 300 and 500 MW and consistently needed at 1,200 MW and above ^[1].

LARS is MISO’s answer, in part, to FERC’s June 18 show cause order in EL26-70-000, which preliminarily found MISO’s tariff lacks study procedures that account for generation serving electrically proximate large loads ^[1]. It sits beside two other pending MISO filings: the Zero Injection Generator Interconnection Agreement in ER26-3552 and the large-load Interconnection Reliability Requirement in ER26-3650 ^[1].

Where the upgrade costs land

The process sorts upgrades into three groups. Problems that show up only in the generation study become generator interconnection projects, and the generator pays ^[1]. Problems found in the load study, or in both studies, become Baseline Reliability Projects or “Other” projects ^[1]. Their costs flow under existing tariff rules, which Google says means recovery from all network customers in the transmission pricing zone ^[3].

There is a second route. LARS models include only queued generators with signed interconnection agreements, so LARS and MISO’s main queue, the Definitive Planning Phase, could count the same transmission headroom twice. MISO calls this “overallocation” ^[1]. If it happens, MISO wrote, the fix becomes a future MTEP project whose cost is “likely to be the load within the transmission pricing zone where the transmission upgrade is required” ^[1].

MISO’s response is a separate filing. It says a pro forma cost shifting agreement and new transmission service terms for large loads will be filed under Section 205 by November 16, and that LARS “need not await these future filings” ^[1].

The exit terms carry their own cost risk. If either side walks away after the study, deposits are forfeited and the parties owe their assigned upgrades ^[1]. If the load leaves within 15 years, the generator gets up to a year, longer if it enters the queue, to obtain new service ^[1]. The clean energy advocates point to what happens after year 15: the generator keeps service of up to 150% of the departed load’s peak “with no further study and no cost at all” ^[7].

Where the sides stand

For it. The MISO Transmission Owners, a group that runs from Ameren, Entergy and Xcel’s operating companies to a long list of co-ops and municipal agencies, call LARS “narrowly tailored” and say it keeps LARS requests out of queue studies ^[6]. The Mississippi and Louisiana commissions call it “fuel agnostic” and say it does not hurt queue participants ^[5].

The Data Center Coalition supports approval and says the 120-day study gives large loads a reason to “proactively contract with new generation supply” ^[4]. It says cost allocation for LARS upgrades “remains an important issue” for MISO’s stakeholder process ^[4]. Its one objection is access: because only the utility can apply, independent generators can’t get to the faster study without the utility, and the coalition asks FERC to encourage MISO to let paired third-party generators apply directly ^[4].

Against it, in part. Google says the concept is one it “continues to strongly support” ^[3]. Its complaint is timing: “MISO has decoupled LARS’s speed to power provisions from cost shifting protections” ^[3]. Under LARS as filed, Google says, customers like it could be kept from using the process without violating their Ratepayer Protection Pledge commitments to pay for the grid upgrades their data centers need ^[3]. It asks FERC to reject the filing with guidance or to accept it, suspend it and consolidate it with the show cause case until the cost agreement arrives ^[3]. Google also objects to the utility gatekeeper, noting that vertically integrated utilities have “a structural incentive to favor their generators” ^[3]. For that pledge it cites its own protest in PJM’s large-load docket, the case whose comment round this column covered in September ^[3].

Against it, outright. Sierra Club, NRDC and four other clean energy groups, joined by two community organizations, say LARS “will shift costs from large loads to other existing and future transmission customers and resource owners” and ask FERC to reject it or issue a deficiency letter ^[7]. They set the filing against MISO’s recent planning cycle: 89 expedited-review requests approved in MTEP26 accounted for roughly 54% of about $14.7 billion in transmission approved, by their reading of a MISO presentation ^[7].

The American Clean Power Association and the Southern Renewable Energy Association compare LARS to MISO’s expedited resource study, ERAS, which FERC accepted only after MISO capped it at 68 requests and set it to expire by August 31, 2027 ^[8]. “It is permanent and uncapped,” they write of LARS ^[8]. They also note that FERC’s show cause order gave “no more than two substations away” as its example of electrically proximate load, while LARS pairs anywhere in a Local Resource Zone ^[8]. MISO considered tighter limits and chose the zone so that landlocked loads can reach enough generation ^[1].

What I’m watching

November 16: MISO’s stated deadline for the pro forma cost shifting agreement ^[1]. Google and the clean energy groups both tied their positions to it ^[3,7]. Whether it holds zone customers harmless for upgrades a paired load and generator cause is the question FERC will have in front of it.

December 2: the effective date MISO asked for ^[1]. FERC can accept, reject or suspend. Whether it waits for the cost filing is open.