The Fuss and Advantages of Siting Large Consumers at Power Plants
Historically, power plants and electric consumers have been miles apart—both figuratively and literally. However, some new power consumers are developing on-site generation or locating their operations at the site of an existing power plant (known as “co-location”). This year, data centers began pursuing co-location with existing nuclear power plants as demand for electricity grew, which immediately triggered regulatory controversy. Now the issue has attracted congressional attention, a dedicated conference before the Federal Energy Regulatory Commission (FERC), and inquiries from state officials.
Several factors are driving co-location. The first two are traditional: cost and reliability. Power demand is surging while regional supply infrastructure is delayed and costly, especially because of red tape. Co-location is a rational response to scarcity on centralized regional power grids.
A novel driver of co-location is that existing nuclear power is well suited to meet tech companies’ ambitious sustainability targets. Nuclear provides consistent carbon-free power, and consuming it on-site ensures verifiable clean energy supply. For environmental stakeholders, this has a credibility advantage over off-site clean energy supply arrangements because verifying indirect emissions remains a work in progress. Long-term purchases from nuclear plants also support the need for improved relicensing policy and make a compelling case for removing costly subsidies (“zero emission credits”), whose value is now unquestionably remunerated via corporate customer dollars.
Co-location trends epitomize the advantages of electric competition. Efficient wholesale markets signal when and where the power system has slack, while retail competition gives consumers access to wholesale markets and reduces barriers to self-supply. Competitive power plant owners are out front, consistent with their incentive to respond to customer preferences creatively. Monopoly utilities are exploring lesser engagements, but their outlook is less promising because their incentive is to expand their regulated capital base, not to respond to customers cost-effectively.
Co-located facilities are not isolated power networks at this stage, given that their host generators remain connected to the central grid. Thus, excitement for pure private grids with no government intervention, while theoretically enticing, is premature. Navigating regulation is inevitable, and proceedings are underway.
A key regulatory complaint is that co-locating demand results in reduced generation deliveries from a plant to the centralized regional market. Generally, this is a red herring. If a farmer sells tomatoes to their neighbors directly rather than through the grocery store, the amount of reduced supply and demand in the central market remains the same. Oftentimes, bilateral arrangements provide customized services in ways that standardized services in a centralized market cannot. Bilateral markets are less transparent than centralized ones, but if consumers want to contract in bilateral markets voluntarily, then there is no consumer welfare argument to deny them that ability.
Rather than worry if new power demand will be met in the central market or the bilateral market, it is better to determine where to site inevitable new large loads. Co-location places demand where there is surplus supply, reducing infrastructure costs like transmission upgrades on the central grid as opposed to developing new demand elsewhere. In other words, enabling bilateral arrangements between generators and customers—a market in itself—can reduce burdens on the central market.
The regulatory controversy does not end there. Co-located facilities can export extra on-site power to the central grid or import power if their demand exceeds on-site supply. This triggers regulatory rules governing power sales on the central grid and allocating shared network costs, such as reserves and transmission. New co-location arrangements are not the first time regulations had to adjust for consumers who partially self-supply and sometimes put power back onto the grid. Such considerations have existed for decades with industrial cogeneration and, more recently, with the broad rise of distributed energy, except instead of determining regulatory rules to compensate household rooftop solar, regulators are now dealing with reactor-sized demand.
Co-location is emerging in many forms, each with fundamentally different implications for the central grid. For example, one tech company told us they might utilize other on-site backup generation or curtail their consumption if the co-located power plant trips offline. Another company said they would maintain full consumption and draw backup power from the central grid. That is a huge difference.
Further, some large consumers may connect behind the meter, while others may connect in front of the meter in a localized “industrial park.” In fact, enabling voluntary industrial parks may be the best hope for any cost-efficient new nuclear revival. After all, it will command a hefty premium that some consumers (like sustainability-minded manufacturers) may willingly shoulder, making it a far better fit for consumer choice policy than socializing marked-up nuclear across all captive ratepayers. Innovative, customer-sited resources are not limited to nuclear, of course, and large consumers seek wholesale and retail competition to take advantage of various self-supply opportunities. For example, industrial consumers are looking into thermal batteries, but the “most critical barriers” to their deployment are lack of access to wholesale power prices and electric rate structures reflective of the underlying costs of delivering power, also known as “cost causation.”
Altogether, different co-location configurations carry various cost implications for the central grid. Regulatory clarity on issues like backup power rules would help. They also cannot be one-size-fits-all, given the heterogeneity of co-located demand.
Indeed, there are valid regulatory considerations at play. However, the proper regulatory response is not to arrest efficient commercial activity, but to update rules to ensure rates reflect the cost that co-location facilities pose to the central grid. Proactive policy minimizes ad hoc future disputes, especially legal challenges and economic complaints that customer self-supply is motivated by inefficiently bypassing the central system. This would help unlock hesitant capital that remains deterred by regulatory uncertainty. That is quite different from efforts by monopoly utilities to throw up regulatory roadblocks, which would chill investment.
FERC will tee up many of these regulatory issues in its forthcoming technical conference on co-location, but whether the commission will unlock market-driven co-location or erect barriers to entry remains to be seen. Interestingly, FERC elected to emphasize state voices at the conference. It would be productive for stakeholders to recognize that the benefits of efficient co-location require retail competition under state purview and wholesale competition under FERC jurisdiction. Confirming the market advantage is a prerequisite to discussion of the nuanced market rules that ensure co-location behavior is efficient and fair to all parties.