Every day, utilities are responding to enormous energy demand.

In July 2026, hourly peak load in the Electric Reliability Council of Texas (ERCOT) reached 91.1 gigawatts (GW), surpassing the previous record in 2023 by 6%. Natural gas generation supplied nearly half of the electricity needed during that peak.

This demand will only continue to grow as more states add large commercial and industrial loads, including data centers. As these campuses push power requirements into the hundreds of megawatts, utilities and developers face a more complex question than where the power will come from: What generation strategy can provide the capacity, reliability and operating flexibility these loads require?

Reciprocating internal combustion engines (RICE) are one behind-the-meter (BTM) option receiving increased consideration, alongside gas turbines, small modular reactors (SMRs), and other power resources. Designing the right generation architecture requires teams to look beyond individual equipment specifications to understand how the generation plant will perform within the broader power system.

With experience spanning power generation, power delivery, energy storage and mission-critical infrastructure, ENERCON helps utilities and data center developers evaluate these technologies as part of an integrated power strategy: one that aligns with the requirements of the load, the realities of the site and how demand may evolve over time.

Why Reciprocating Generation Is Getting a Closer Look

Historically, RICE have accounted for a very small share of natural gas power plants. Compared to natural gas combustion turbines and combine-cycle units, RICE have considerably less capacity and have been considered better suited for backup or emergency power. The renewed interest in larger utility-scale power generation applications is driven in part by advancements in RICE technology, the sustained need for more power, and changes in natural gas markets that provide relatively stable fuel prices.

In recent years, larger RICE units have been installed throughout the US. These units can be installed at a single generating facility to bring the total plant capacity to 200 MW or more. And while capacity, efficiency and cost remain important factors, the right technology also depends on how the system will operate.

How quickly does capacity need to become available? Will demand be added in phases? What happens when a generating unit unexpectedly goes offline? What site, fuel supply and environmental constraints need to be addressed?

Those questions can make the modular nature of reciprocating generation enticing.

Rather than concentrating a large portion of plant output in a few generating units, a RICE plant distributes capacity across multiple engines. If one unit is unavailable because of a trip or planned maintenance, the event removes a smaller portion of the plant’s total capacity. That can give operators greater flexibility in maintaining reserve capacity while continuing to serve the load.

Modularity can also support phased development. Additional generating units can potentially be added as demand grows rather than requiring the full buildout capacity to be installed all at once.

Reciprocating engines also offer fast-start capabilities and can perform efficiently across a range of loads. These characteristics become more valuable when generation needs to respond flexibly to changing operating conditions.

At the same time, those advantages come with tradeoffs. A large reciprocating plant requires more individual generating units and associated equipment. Site footprint, maintenance requirements, controls, switchgear, emissions, noise and natural gas infrastructure all factor into the evaluation.

RICE vs. Gas Turbines: Measuring Beyond Megawatts

Gas turbines offer a different approach to solving the same capacity challenge.

A turbine can provide a much larger block of power from a single unit, reducing the number of generators and associated systems required for a large plant. Turbines can also be attractive where generation is expected to operate continuously at high output or where site constraints favor a more concentrated generation footprint.

The tradeoff here is that larger generating units create larger individual contingencies. When a large turbine goes offline, the system can lose substantially more capacity in a single event than it would with the loss of one reciprocating engine. That distinction affects how reserve capacity and the surrounding electrical system are designed.

It also illustrates why there is no simple megawatt threshold where one technology automatically becomes the better choice. Total capacity matters, but so do the operating profile, site constraints, natural gas supply, environmental requirements, project schedule, maintenance strategy and long-term economics.

Comparative Assessment

Reliability as a System-Level Decision

For a data center, small differences in availability can have significant consequences. A target of 99.999% availability—often referred to as “five nines”—equates to roughly five minutes of potential downtime over an entire year. Meeting that level of performance requires the entire power system to be designed around how it responds when something goes wrong.

That design often starts with a redundancy model. In an N configuration, the system has the capacity required to serve the design load, but no additional capacity if a critical component becomes unavailable. N+1 adds one redundant component or unit of capacity, allowing the system to maintain its required output following a single equipment loss. A 2N architecture provides two independent systems, each capable of supporting the full design load. More robust configurations can add redundancy beyond those basic models depending on the criticality of the facility and the owner’s reliability objectives.

Reciprocating generation can fit particularly well within an N+1 strategy because capacity is distributed across multiple smaller units. If one engine trips or is taken offline for maintenance, the remaining units and reserve capacity can be designed to continue serving the required load. The smaller size of each generating block also reduces the magnitude of the contingency the rest of the system must absorb.

Redundant generation, however, does not eliminate the transition between the loss of one resource and the response of another. Even when reserve generation is available, it takes time for an offline unit to start, synchronize and accept load.

This is where battery energy storage systems (BESS) become an integral part of the reliability architecture. Because batteries can respond in milliseconds, they can provide immediate support following a generation loss while available generation responds or reserve units come online. Once sufficient generation is available, the BESS can transition back toward its planned operating state and the system can restore its intended level of redundancy.

The same redundancy philosophy extends beyond generation. Depending on the facility’s requirements, the batteries, UPS systems, switchgear and electrical distribution may incorporate N+1, 2N or other redundant configurations so that the failure of a single component does not create a direct path to the critical load.

For data centers pursuing five-nines availability, the objective is not simply to maximize the number of redundant components. It is to establish an appropriate redundancy model and design each layer of the power system to respond together when a credible failure occurs.

Designing for the Load Today and the System Tomorrow

Large data center loads are often developed in phases, making the long-term generation strategy as important as the initial plant.

A reciprocating plant’s modularity may allow capacity to be added incrementally as demand materializes. A turbine-based solution may offer different advantages as the required capacity grows. Future utility infrastructure could also change how local generation is operated over time.

Decisions about natural gas supply, electrical infrastructure, site layout and environmental permitting for the first phase can influence what is practical later. Evaluating the ultimate development scenario alongside immediate capacity needs can help utilities and developers avoid choices that unnecessarily constrain future expansion.

Ultimately, the question is not simply whether reciprocating engines or gas turbines are better suited for data center growth. It is which generation architecture best aligns with the way a particular power system needs to operate.

ENERCON brings experience across the entire power lifecycle to help clients evaluate these tradeoffs as an integrated system. By considering generation technology alongside reliability requirements, site conditions and future capacity needs, utilities and data center developers can make more informed decisions about the infrastructure needed to support continued growth. Learn more about ENERCON’s power generation, power delivery and grid infrastructure capabilities.