The Potential Of Multi-Fuel Novalt™16 Gas Turbine For Data Center Application

Abstract

The power demand growth from the data center market is boosting the need for gas turbines at the point that manufacturers are ramping up their capacity, struggling to meet such a large and urgent request. In this context, the performance of the NovaLT™16 gas turbine (GT) is assessed under a prescribed daily electricity load profile, assumed as a sequence of steady-state operating points. The load shape is defined considering the challenges posed by artificial intelligence (AI): if inference overtakes training as the dominant load source, the daily peak will depend on the type and time of usage. Accordingly, a low training - high business utilization scenario is taken into account on two representative hot and cold days (based on average daily temperature in July in Phoenix and January in Chicago, respectively).

Given its ability to start and run on different types of fuel, natural gas (NG) is completely replaced by two sustainable options: hydrogen (H2) and hydrotreated vegetable oil (HVO). The focus is on identifying the most efficient GT control strategy in load-following mode, which is significantly influenced by fuel properties, with a direct impact on NOx emissions. In addition, carbon dioxide savings compared to conventional fuel are estimated.