Can Hydrogen Help Decarbonize Your Turbine Fleet?

Unfortunately, today, a 75% blend is impractical. That may change in the next few years as huge investments are made (driven partly by the Inflation Reduction Act), but it won’t be easy. A recent pilot project by New York Power Authority (NYPA) and GE did reach up to 44% hydrogen with a GE LM6000, but commercial operations at that level were deemed unsustainable for now (read the executive summary here). Their actual CO₂ readings were consistent with our graph: 35% hydrogen reduced emissions by ~14%. While the technical challenges are likely surmountable, the main concern was hydrogen availability. Maintaining a stable flame at a 35% blend requires high H₂ volumes and consistent delivery pressures. There isn’t enough green hydrogen to meet that demand in many places. Help may be on the horizon as new Department of Energy-supported hydrogen hubs are developed nationwide.

So, how much hydrogen is needed to support a blend ratio of 35% for a peaker plant? How would we produce and store it? Does it need to be “green” hydrogen, or should we consider other hydrogen sources? It’s a complex problem that requires some advanced techno-economic modeling.

One approach is to run a detailed market simulation of the generator to find its optimal schedule without any hydrogen resource constraints. After all, the market doesn’t care what fuel a generator uses, only its cost vs. market prices. Our fuel price must account for the MMBtu blend ratio of natural gas and hydrogen. Monte Carlo simulation is useful here to account for risks if it produces operational results for each scenario (like PCI’s GenTrader Stochastic module) that can be used to determine hydrogen needs. You can use historical prices and actual weather data to ensure a realistic relationship between Locational Marginal Prices (LMPs), ancillary service prices, gas prices, and renewable availability. Past prices don’t guarantee future performance, but help provide realistic scenarios. If you have a trustworthy long-term price and renewable forecasts, that’s fine too.

Once you have an optimal generation schedule for a year (or more), you need another model to simulate hydrogen production and storage for your desired blend ratio. Be sure to use the same scenarios for optimal generation and hydrogen analysis.

The hydrogen model should make it easy to experiment with the dozens of parameters that affect hydrogen production, storage, and delivery to determine the best way to meet optimal generation demand. There are several ways to produce hydrogen with varying degrees of carbon intensity, operational complexity, investment cost, and risk.

Here are a few things to ask, test, and consider:

• Can you realistically meet demand with renewable electrolysis?
• What size and efficiency electrolyzer do you need? Does it need to provide grid services?
• Is your renewable Capacity Factor (CF) enough for electrolysis, or would you be better off using grid power and buying renewable energy certificates (RECs)?
• What about steam-methane reforming (SMR) with carbon capture or offsets?
• Given the anticipated generation blend and schedule, what size storage do you need?
• Can you sell excess produced hydrogen at a profit? Where? How?
• What effect do hydrogen and gas prices have on my net present value (NPV)?

PCI has developed the Techno-Economic Modeling and Analysis (TEAM) analytics, including sensitivity analysis, to help answer these questions.

The following graphs illustrate electrolysis hydrogen production, storage, and consumption by a CCGT with a 30% blend. It shows that the H₂ storage capacity is between 5,000 and 42,000 kilograms. Excess production is sold with an offtake agreement.