Oklahoma Corporation Commissioners are being advised by their own Public Utility Division to take certain steps to prepare for an expected flood of demands for increased electricity generation in the coming few years.
It’s not a matter of 15 to 20 years down the road but perhaps 3 to 5 years.
As Oklahoma confronts a growing demand for more and more electricity caused in part by industrial growth across the state, state regulators are mulling how to deal with it, not just from a matter of providing electricity but also in dealing with the cost of it, from the utility standpoint and that of ratepayers.
The Corporation Commission’s Public Utility Division recently presented its case during an exploratory discussion of the topic held last week involving the Southwest Power Pool, Oklahoma Gas and Electricity and Public Service Company of Oklahoma.
“Oklahoma is experiencing rising electricity demand from large-scale data centers and AI infrastructure,” stated the PUD in a report to the commissioners.
The recommendations came from Mark Argenbright, director of PUD and Deputy Director Fairo Mitchell.
“While this research focuses on data center and AI driven load growth, Oklahoma utilities are also seeing large loads from manufacturing, refineries, federal entities, and crypto currency mining that may require the attention of the OCC as well,” explained Argenbright.
III. Recommended Options for the OCC
1. Require third-party validation of large load forecasts in high-load applications.
2. Require utility to provide the stage of development e.g. inquiry or under contract.
3. Require the utility to identify that contribution in aid of contraction (“CIAC”) has been established or paid.
4. Mandate that data center infrastructure costs be ring-fenced or directly assigned to beneficiaries.
5. Mandate that all infrastructure costs be verified by a third party.
6. Mandate minimum contract time frames and exit fees if contract time is not met to mitigate stranded investment and ratepayers’ subsidization.
7. Require standardized tariffs for high-load customers to replace the need to use special contracts.
8. If special contracts are used, ensure they are in alignment with long-term planning and do not shift risk allocation.
9. Inquire about tax incentives and grid impact.
Below is the entire Executive Summary:
Oklahoma is experiencing rising electricity demand from large-scale data centers and AI infrastructure.
These loads, typically exceeding 100 MW per site, are unique in their scale, consistency, and demand for high reliability. While touted as offering economic development opportunities, these opportunities appear to be limited because they do not require many employees, and these facilities will not spur other facilities based on their existence unlike other large electricity loads such as a car manufacturing facility. They also raise regulatory concerns related to grid reliability, ratepayer fairness, and prudent infrastructure investment.
This research outlines key regulatory risks and offers actionable recommendations for the Oklahoma Corporation Commission (“OCC”) when considering large load. It also summarizes how other states are proactively managing similar challenges. The OCC has a unique opportunity to implement policies that
ensure cost transparency, prevent cross-subsidization, and align electric utility investments with longterm state goals.
I. Purpose
This research provides the OCC with a regulatory framework to assess and manage the implications of data center and artificial intelligence (“AI”) driven load growth. It identifies current risks to grid reliability and rate structures, reviews policy actions in other states, and recommends regulatory strategies to protect Oklahoma ratepayers and ensure responsible utility planning.
While this research focuses on data center and AI driven load growth, Oklahoma utilities are also seeing large loads from manufacturing, refineries, federal entities, and crypto currency mining that may require the attention of the OCC as well.
II. Key Regulatory Issues
1. Load Forecasting and Infrastructure Planning
• Utilities may overestimate demand due to speculative or overlapping data center proposals.
• Unverified projections may be used to justify large capital investments in generation and transmission. Historically only around thirty percent (30%) of generation projects in the
Southwest Power Pool generation interconnection queue are built and come online.
2. Cost Allocation and Ratepayer Fairness
• Utilities may socialize infrastructure costs across all customer classes.
• Residential and small commercial customers may bear increased rates without benefiting from these developments.
• Large loads may curtail usage and avoid demand charges that will be shouldered by other rate classes.
3. Special Contracts and Tariffs
• Discounted or custom rates for data centers may lack transparency and may not cover their marginal costs.
• Inconsistent contract terms complicate planning and can shift risk to other customers.
• Interruptible rates are not favored because operations require extremely high reliability and involve long lead times to shift or curtail load. Therefore, data centers are not viable demandside management tools which may aggravate grid constraints in high congestion times.
• Some states have suggested large loads self-supply power to island them from the grid and to address grid congestion.
Oklahoma has attempted to address this issue by creating a behind the meter statute. This statute eliminates the obligation to serve and alleviates the associated reserve requirement at Southwest Power Pool (“SPP”). However, Oklahoma’s statute does not require the large load to supply its own power. It does allow the large load customer to market to other businesses the rate that it has from the utility. This would appear to aggravate congestion, but not to help to mitigate congestion and further complicates cost shifting issues.
A barrier to this self-supply option is that the data center is not in the power production business. Hence the data center may look to the local utility to oversee the construction and operation of the energy producing facility for a contracted period with the local utility purchasing the unit(s) at an agreed upon price at the end of the contract (e.g. 20 years).
4. Risk of Stranded Assets
• Infrastructure built for short-term or mobile data center loads may become underutilized.
• Large load customers may exit before full cost recovery, leaving ratepayers with stranded costs.
5. Reliability
• Maintaining a supply and demand balance on the grid is of upmost importance as utilities in Oklahoma have an obligation to serve. This obligation also comes with FERC mandated planning
reserve margins from SPP.
• Data centers/large loads could switch offline and create a load shedding event in the SPP regional transmission system (“RTO”).
o Increased RTO transmission expenses could also be passed on to other customers from increased pricing during load shedding events.
III. Recommended Options for the OCC
1. Require third-party validation of large load forecasts in high-load applications.
2. Require utility to provide the stage of development e.g. inquiry or under contract.
3. Require the utility to identify that contribution in aid of contraction (“CIAC”) has been established or paid.
4. Mandate that data center infrastructure costs be ring-fenced or directly assigned to beneficiaries.
5. Mandate that all infrastructure costs be verified by a third party.
6. Mandate minimum contract time frames and exit fees if contract time is not met to mitigate stranded investment and ratepayers’ subsidization.
7. Require standardized tariffs for high-load customers to replace the need to use special contracts.
8. If special contracts are used, ensure they are in alignment with long-term planning and do not shift risk allocation.
9. Inquire about tax incentives and grid impact.
IV. Actions Taken by Other States
State Key Action Implication for Oklahoma
Arizona Allows Small Modular Reactors (“SMR’s”) without
full environmental review. Alternative generation flexibility.
California Requires full cost recovery from data centers.
Cost containment.
Kentucky Limits special contracts and mandates fixed cost recovery. Ratepayer protection.
Virginia Legislation to evaluate data center subsidies. Proactive subsidy oversight.
Texas Transmission charges shared even with behind-the-meter
generation. Equity issue in recovery.
Georgia Allows minimum usage contracts and long terms. Supports utility cost planning.
Ohio Implements structured data center tariff. Transparency and
consistency in pricing.
Oregon Requires 10-year minimum usage contracts. Discourages speculative load.
Click here to view public meeting