The divergence between these two versions is wide enough that the final text — if a bill passes — could look materially different from either current draft
The System Pressure
Michigan Democrats introduced a coordinated multi-bill legislative package in mid-2026 that, if enacted, would structurally reorder the terms under which data centers access power, water, and permits in the state. The central mechanism is an interconnection gate: under HB 6137 and SB 1050, the Michigan Public Service Commission (PSC) would be prohibited from approving any data center interconnection until a community benefit agreement has received local government approval in the relevant jurisdiction. Site plan approvals and construction permits would be blocked on the same condition.
For operators already managing interconnection timelines that run three to seven years in most markets, adding a mandatory community negotiation phase before the PSC process begins introduces a new upstream constraint with no defined clock. That sequencing risk sits directly in the critical path of new capacity.
This is not a single-issue package. The bills span energy tariffs, clean energy procurement, water permitting, decommissioning obligations, cooling system design, labor standards, and NDA restrictions. Each element changes a cost or process variable independently. In combination, they define a new threshold for what it costs and takes to operate in Michigan.
The Drivers, Dependencies, and Constraints
The community benefit agreement framework (HB 6137, SB 1050) requires operators to address at least half of five specified topics: local hiring, water use, infrastructure cost allocation, community investment funding, and home efficiency improvements. The agreements are legally binding contracts between operators and local governments, utility providers, or other community stakeholders. Six other states have considered similar CBA requirements for data centers in 2026, but the explicit linkage to PSC interconnection approval gives Michigan’s version structural teeth that advisory or voluntary frameworks lack.
On energy, the Senate and House have produced competing drafts with a material gap between them. Senate Bill 1047 requires data centers to procure clean energy covering 90% of annual usage, execute minimum 20-year utility contracts with 90% minimum billing of contract demand, pay a $100,000 application fee, participate in demand response programs, and absorb their own decommissioning costs. House Bill 6135 takes a structurally different approach: it creates a large-load commercial rate class with 15-year minimum contracts and places energy generation costs on operators, but imposes no clean energy mandate and carries fewer financial conditions. The divergence between these two versions is wide enough that the final text — if a bill passes — could look materially different from either current draft.
Water constraints introduce a third dimension. Senate Bill 1046 would require permits for any entity using 550,000 gallons or more per day and cap consumptive use at 2 million gallons per day for permit holders. Large data centers can consume up to 5 million gallons per day under current evaporative cooling models, meaning the 2-million-gallon cap would require significant cooling infrastructure changes for high-density facilities. The House equivalent is a design-level constraint: new data centers would be required to use closed-loop cooling systems or source water exclusively from municipal systems — a requirement that shapes capital expenditure from the foundation stage. Twelve states have considered closed-loop cooling legislation in 2026, making this a national trend rather than a Michigan peculiarity.
Decommissioning requirements in HB 6142 would prohibit PSC approval of data center contracts unless a decommissioning plan with financial assurance is included, subject to mandatory PSC review every three years. Decommissioning legislation has been introduced in twelve states this year. Labor provisions in HB 6141 and SB 1048 would condition PSC contract approval on construction and maintenance work being performed under a collective bargaining or project labor agreement — a procurement constraint for EPC relationships not currently structured to those terms.
Open Dependencies
None of these bills have been enacted as of mid-2026. The Senate and House energy frameworks differ on the two variables that matter most to portfolio economics: clean energy mandate and contract duration. Whether a reconciled version adopts the Senate’s 90% clean energy requirement or the House’s rate class approach is not resolvable from current legislative text.
The enforceability and pace of community benefit agreements depend on local government capacity to negotiate and approve plans. The bills do not define a review timeline for local approval, which means the pre-interconnection gate could produce diffuse, unpredictable delays rather than a defined process. Whether the PSC would have authority to impose CBA review deadlines is not addressed in the current language.
Michigan’s tax incentive landscape is separately unsettled. Legislation has been introduced to both repeal existing incentives and extend exemptions in brownfield zones, with neither resolved. The NDA restrictions in HB 6140 and SB 1049 — which prohibit NDAs on projects receiving tax incentives, with an intellectual property carveout — interact with whichever tax framework survives, creating a compliance dependency that cannot be assessed until both questions are settled.
The Operating Exposure for Global Heads of Data Center Energy
The interconnection gate is the highest-priority item to track. If HB 6137 and SB 1050 pass as introduced, any Michigan facility in active development or site selection must account for a pre-queue community engagement phase that currently has no defined timeline. Projects that have not modeled this sequencing risk may carry materially optimistic commissioning assumptions.
If the Senate’s clean energy mandate survives in final legislation, it would require a procurement strategy adjustment for operators currently relying on standard utility supply or unbundled RECs. A 90% annual clean energy obligation combined with a 20-year contract term represents a longer and more rigid offtake commitment than most large-load utility agreements in current use. Demand response participation as a contract condition adds an operational flexibility constraint on top of the procurement requirement.
Water exposure is a capital planning issue, not only a compliance issue. Facilities designed for evaporative or open-loop cooling face a structural gap if the consumptive use cap or closed-loop requirement is enacted. Retrofitting high-density cooling infrastructure at operating scale is expensive and time-consuming; the constraint applies at the design stage for new builds, making early-stage site assessment the correct intervention point.
Signals the System Is Shifting
The breadth of this package is itself a signal. When a single legislative session introduces coordinated instruments across energy, water, labor, decommissioning, and community benefit simultaneously, the probability of partial enactment is meaningfully higher than for isolated bills. The multi-front structure also increases the chance that operators negotiating one provision will face unexpected obligations on another.
The national pattern reinforces this read. Six states have considered CBAs, twelve have considered decommissioning requirements, and twelve have considered closed-loop cooling mandates in 2026. Michigan’s version is among the most operationally demanding currently in motion, but the direction is consistent with what is accumulating across the country.
The next material signals to watch: whether the Senate’s clean energy and 20-year contract requirements survive conference; whether the PSC interconnection gate passes with or without a defined CBA approval timeline; and whether Michigan resolves its tax incentive question in a way that changes the NDA exposure calculus. A Michigan statute incorporating SB 1047’s energy terms would be the most demanding state energy framework currently applied to large data center loads in the US and would likely accelerate comparable provisions in adjacent state legislatures.
Sources
- Multistate — Michigan Data Center Legislation: Energy, Water & Labor Rules (Link)
