On the Estimation of Exhaust Gas Recirculation Flow and Waste Heat Recovery Tradeoffs Based on Differential Pressure Measurements

Abstract

Both exhaust gas recirculation (EGR) and waste heat recovery (WHR) are attractive technologies for more efficient spark ignition (SI) engines. The fuel economy benefit of cooled external EGR on SI engines is well established, and preliminary first law analysis of engine energy flows indicates the large potential for efficiency improvements with WHR. Nevertheless, both technologies face major challenges that need to be addressed to become viable solutions for more efficient SI engines.

Cooled external EGR improves SI engine efficiency under wide range of conditions. However, inaccurate estimation of the EGR fraction in the intake manifold can be detrimental as it can lead to inaccurate air charge estimation, knock and misfire. Accurate EGR estimation based on a differential pressure (ΔP) measurement is very challenging at the low ΔP 's due to pressure pulsations and inertial effects. While some systems are capable of increasing ΔP across the EGR valve to improve EGR estimation, the higher ΔP is undesirable as it can increase pumping losses. EGR estimation accuracy at low ΔP can be improved by fast sampling of the ΔP signal and using the newly derived approximations of the unsteady compressible flow orifice equation. Both experimental data from a modified Ford 1.6 L EcoBoost engine with added LP and HP-EGR loops, and simulation predictions from its GT-Power models are used to evaluate the estimation methods. A sampling frequency of at least 1 kHz reduces the ΔP lower bound required to keep the LP and HP-EGR estimation error within a target 1% from 12.7 and 27.9 to 1.9 and 2.9 kPa respectively. The LP ΔP lower bound can be further reduced to 1.1 kPa with variable filtering, but the sampling frequency requirement is increased to 3 kHz to achieve the full benefit.

The impact of gauge-line distortions and EGR valve area offset errors are also evaluated. An extension of a preexisting lumped parameter model is proposed to estimate the actual ΔP from the distorted measurement. Simulation results show that the proposed model can correct for the gauge-line errors under modeled pressure measurement noise. Valve area offset errors are shown to have substantial impact on the EGR estimation error especially for the HP-EGR case. A novel online calibration method for the HP-EGR valve area using preexisting engine sensors is developed and shown to have promise for implementation.

The second part of this thesis studies the limitations and challenges of WHR through electric turbo-generation. Insights into the tradeoffs between exhaust energy recovery and increased pumping losses from the flow restriction of the electric turbo-generator (eTG) are provided and assessed using thermodynamic principles and with a detailed GT-Power engine model. The additional pumping losses are load independent and cannot be offset by the eTG power at low loads. Engine simulations are used to predict the influence of the increased back pressure on pumping work, in-cylinder residuals and combustion. The eTG is detrimental at the high loads as it requires more spark retard to mitigate the increased knocking tendency. The eTG benefit is therefore restricted to mid-loads. The reduction in fuel consumption possible over various drive cycles is estimated based on the steady-state efficiency of frequently visited operating points assuming all recovered energy can be reused at a representative engine efficiency. Fuel consumption reductions of 1.2% are projected for the combined cycle.