Navies worldwide are exploring the use of alternative fuels for their vessels, aiming to cut down greenhouse gas emissions and lessen their reliance on fossil fuels. This effort also seeks to reduce their environmental impact and emissions signatures. However, combustion engines running on low reactivity fuels, including most alternative fuels like natural gas or alcohols, are limited by their lower load acceptance compared to diesel engines. As a result, they may not meet the stringent naval requirements for dynamic load capacity in power generation. A high dynamic loading capacity is a precondition for high maneuverability of the vessel on the one hand and handling of pulsed power loads and step loads caused by rail-guns and directed energy weapons on the other hand. Previous research into the dynamic response of natural gas engines required detailed in-cylinder combustion models to predict knock. These 0D/1D simulation models rely on extensive data to calibrate the combustion model, which is generally unavailable to the naval engineer designing a propulsion or energy system. Additionally, these simulation models require a significant amount of computational power and do rarely run in real-time. This study predicts the dynamic response using a Mean Value First Principle (MVFP) engine model based on the filling and emptying approach and turbocharger performance maps derived from limited data and measurements. Initially, the model is calibrated using engine data provided by the manufacturer and experimental measurements in steady state on a spark-ignited Caterpillar 3508A gas engine driving a generator at a constant speed of 1500 rpm. Additional calibrations are performed using experimental measurements of the dynamic response to step loads. Analysis of the simulation results reveals the model’s capability to predict the dynamic response of the air intake and exhaust system while being able to run in real-time. The results further show that steady-state simulations do not consider the effect of turbocharger inertia and lagging fresh air supply on the thermal loading and knock probability of the engine sufficiently. This paper underlines the significance of implementing the air and exhaust path dynamics, including turbocharger performance models, in mean-value models when investigating combustion engines operating on low reactivity fuels. Furthermore, the paper provides guidance on the minimum required number of measurements and load steps to calibrate a mean value model for the prediction of the engine operating parameters under dynamic loads. The resulting model can be used to establish limitations for the loads on the electric grid and evaluate control strategies to improve the combustion engine generator and electrical systems’ performance.