Jay O. Keller

Hybrid energy storage systems for stand-alone electric power systems: optimization of system performance and cost through control strategies

Abstract The development of remote, renewable-based energy is hindered in part by the lack of affordable energy storage. Requiring power-on-demand from an energy system powered by intermittent or seasonal sources may necessitate one-month’s energy storage—an expensive proposition using conventional storage technologies. If multiple energy storage devices with complementary performance characteristics are used together, the resulting ‘hybrid energy-storage system’ can dramatically reduce the cost of energy storage over single storage systems. The coupling of conventional storage batteries with emerging hydrogen technologies provides one such hybrid system. Hydrogen energy storage in this context includes an electrolyzer, hydrogen storage tank, and a fuel cell. An additional component that must be considered is the control system that determines when the various components are used. Since the control system has an effect on component sizes and thus system and operating costs, the control algorithm must be carefully considered for any system with energy storage. For this study, a time-dependent model of a stand-alone, solar powered, battery-hydrogen hybrid energy storage system was developed to investigate energy storage options for cases where supply and demand of energy are not well matched daily or seasonally. Simulations were performed for residential use with measured solar fluxes and simulated hourly loads for a site at Yuma, Arizona, USA, a desert climate at 32.7 N latitude. Renewable-based power not needed to satisfy the load is stored for later use. Two hybrid energy-storage algorithms were considered. The first is a conventional ‘state-of-charge’ control system that uses the current state of the storage system for control. The second control system presumes knowledge of future demand through a feed-forward, neural net or other ‘intelligent’ control systems. Both algorithms use battery storage to provide much of the daily energy shifting and hydrogen to provide seasonal energy shifting, thus using each storage technology to its best advantage. The cost of storing energy with a hybrid energy-storage scheme was found to be much less expensive than either single storage method, with a hybrid system storage costing 48% of the cost of a hydrogen-only system and only 9% of the cost of a conventional, battery-only system. In addition, the neural-net control system is compared to a standard battery state-of-charge control scheme, and it is shown that neural-net control systems better utilize expensive components and result in less expensive electric power than state-of-charge control systems.

Thermodynamic analysis of hydrogen production by steam reforming

Abstract This paper presents thermodynamic analysis of hydrogen production by steam reforming. The analysis treats the chemistry at two levels: a global species balance assuming complete reaction and solution of the equilibrium composition at the specified reformer temperature. The global reaction allows for an energy balance that leads to analytical expressions for the thermal efficiency. We use this to determine the maximum efficiency, and to distinguish between various definitions of efficiency. To obtain a more realistic estimate of the efficiency, the chemical equilibrium solution is combined with a system energy balance, which compares the energy required to vaporize and heat the fuel–steam mixture to the reformer temperature with the heat available from combusting the residual fuels in the reformate stream. The equilibrium solutions are compared to experimental measurements of the species and thermal efficiency of reforming diesel fuel, obtained with prototype compact steam reformers. The observed efficiency is significantly lower than the equilibrium prediction, indicating that both incomplete reaction and heat transfer losses reduce the performance.

Pulse combustion: The importance of characteristic times

Abstract The response of a valved pulse combustor to changes in the relative timing between the resonant pressure wave and the instantaneous energy release rate has been examined. Experiments were designed to examine the pulse combustors response to independent changes in the experimental conditions that resulted in nearly independent changes in the fluid dynamic species mixing time, the fluid dynamic mixing time of cold reactants with hot products, the characteristic chemical kinetics time, and the characteristic resonance time. The time scales considered in this study were adjusted independently to modify the coupling between the instantaneous energy release rate and the resonant pressure wave, thereby modifying the magnitude of the pressure oscillations and altering the frequency of operation. All of these experimental observations of the pulse combustor response to variations in characteristic time scales are interpreted in terms of Rayleighs criterion.

Thermodynamic comparison of fuel cells to the Carnot cycle

Abstract This paper compares the theoretical maximum efficiency of a fuel cell to the efficiency of a Carnot cycle driven by the same net reaction. The comparison dispels the misconception that an ideal fuel cell is potentially more efficient than an ideal heat engine. The paper presents expressions for the thermal efficiencies of an ideal fuel cell and a Carnot heat engine. To show that the maximum efficiency is the same, the analysis of the Carnot cycle is modified to consider an engine driven by a combustion reaction. The derivation invokes the approximations that the enthalpy and entropy changes for the reaction are independent of temperature; these approximations are justified by the hydrogen-oxidation reaction. The analysis extends that presented by Appleby and Foulkes (Fuel Cell Handbook, Van Nostrand Reinhold, New York 1989) by showing that with proper accounting for heat addition, the maximum efficiency of a fuel cell is 100%—not larger—for reactions with a positive entropy change. In addition, this paper explains the difference between the combustion temperature, at which an idealized Carnot cycle would operate, and the adiabatic flame temperature.

A hydrogen fuelled internal combustion engine designed for single speed/power operation

Abstract Sandia National Laboratory is developing from first principles a hydrogen fuelled internal combustion engine for driving an electrical generator that can be utilized either as a stationary power set or the auxiliary power unit in a hybrid vehicle. The intent is to take advantage of hydrogens unique fuel characteristics and the constant speed characteristics of generator sets to maximize thermal efficiency while minimizing emissions. The current experiments utilize a flat cylinder combustion chamber shape with two ignition points at high (14:1) compression ratio. Emissions and indicated thermal efficiency measurements with fuels of hydrogen, natural gas and a blend confirm low emissions and high thermal efficiency. CFD modelling done by Los Alamos National Laboratory (Los Alamos, NM) using their KIVA code is helping to further direct variations in the experimental design space.

Pulse combustion: The mechanisms of NOx production☆

Abstract A study has been performed to elucidate the fundamental mechanisms controlling the production of NO in a Helmholtz-type pulse combustor. Cycle-resolved measurements of velocity and temperature in the combustion chamber were made. These measurements were combined with the Zeldovich NO formation mechanism to explain the mechanism responsible for the low NO formation found in pulse combustors. The mechanism responsible for low NO formation found in pulsating flow as compared to nonpulsating flow was found to be a short residence time at high temperature. Three different possible mechanisms for this short residence time were investigated: (1) hot products from combustion are cooled by mixing with the cooler exhaust gases entering the combustion chamber from the tail pipe, thus quenching the NO formation reaction (“Automatic Exhaust Gas Recirculation”), (2) hot combustion products are quickly cooled by mixing with the incoming cold reactants, and (3) residual products with a lower overall temperature (due to an increased rate of heat transfer in the combustion chamber) readily mix with hot products producing a short residence time at high temperature. It is shown that the mechanism responsible for the low NO emission in pulse combustors is a result of item 3. A short residence time at high temperature is caused by rapid mixing with cooler residual gases that are lower in temperature due to increased rates of heat transfer in the combustion chamber.

Pulse combustor tail-pipe heat-transfer dependence on frequency, amplitude, and mean flow rate

A commonly cited advantage of pulse combustors is a high rate of heat transfer in the tail pipe. Past research on these rates of heat transfer is inconclusive regarding the amount of heat transfer enhancement and how various flow parameters affect this enhancement. This article reports an experimental heat transfer study in the tail pipe of a pulse combustor. The pulsation frequency, pulsation amplitude, and mean flow rate were varied systematically, and their effects on the heat transfer rates assessed. Spatially averaged Nusselt numbers were obtained from thermocouple measurements using a standard log-mean heat exchanger calculation. The Nusselt number was found to increase with both pulsation amplitude and frequency, with a maximum enhancement of 2.5 times that of steady flow at the same mean Reynolds number. The Nusselt number enhancement decreased with increasing mass flow rate for a given combustor pulsation frequency and amplitude. Independent axially resolved heat flux and gas temperature measurements confirmed the large Nusselt number increase with pulsations and demonstrated that entrance effects, although present, were small compared to the Nusselt number enhancement due to the pulsations. The data are compared with quasi-steady theory, which is the only available theory in the literature for this problem. Quasi-steady theory does not account for frequency effects and is not adequate for describing the data from this study.

Heat transfer enhancement in the oscillating turbulent flow of a pulse combustor tail pipe

Abstract Heat transfer rates in pulse combustor tail pipes and in other reversing, oscillating, turbulent flows have been found to be much higher than those of steady turbulent flow. To elucidate the mechanisms of the enhancement, the temperature and velocity fields, measured with two-line atomic fluorescence (TLAF) and laser Doppier velocimetry (LDV), respectively, are compared. Time-resolved wall heat fluxes and Nusselt numbers are also presented and discussed. Possible causes for the heat transfer enhancement in oscillating flows are reviewed and discussed in view of the data presented in this paper and the recent literature.

Time-resolved velocities and turbulence in the oscillating flow of a pulse combustor tail pipe

Abstract The cyclic behavior of the oscillating velocity field in the tail pipe of a pulse combustor was studied using laser doppler velocimetry. In this flow, the oscillations result from an acoustic resonance and have amplitudes of up to 5 times the mean velocity. Oscillation frequencies were varied from 67 to 101 Hz. Streamwise velocity and turbulence-intensity boundary layer profiles were measured to within 130 μm of the wall, and transverse turbulence measurements were made to within 2 mm. The phase relationships of the velocity, turbulence intensity, and combustion chamber pressure oscillations are compared. Velocity oscillations near the wall are found to phase lead those in the center of the pipe, creating periodic flow reversals through the boundary layer. A comparison is made between this turbulent oscillating boundary layer and the laminar oscillating boundary layer for flow over a flat plate. The effects of axial position, pulsation frequency, pulsation amplitude, and mean flow rate on the velocity and turbulence profiles are discussed. Time-resolved wall shear stresses (directly calculated from the velocity measurements) are presented and compared with those of steady turbulent flow. Time-averaged velocity and turbulence profiles are also compared with those of conventional steady turbulent flows. The time-averaged velocity profile is found to be flatter than that of steady flow at the same mean Reynolds number, and both the streamwise and transverse turbulence intensities are found to be significantly higher than those of steady flow.

NOx and CO emissions from a pulse combustor operating in a lean premixed mode

Abstract Control over the combustion fluid dynamics was used to minimize the emission of NO x and CO. The combustion kinetics were controlled by operating the combustor premixed and were varied by modifying the equivalence ratio over the lean stability envelope. A wide dynamic range in fluid dynamic mixing characteristics was also investigated by modifying the degree of macroscopic mixing and microscopic mixing. The residence time at high temperature was controlled by modifying the frequency of the periodic reacting flow in a pulse combustor. It was found that controlling the flame temperature, chemical kinetics, and residence time at high temperature was best accomplished by controlling the equivalence ratio and the degree of macroscopic mixing rather than controlling the microscopic mixing over the dynamic range obtainable by the techniques used in this study. Emission levels below 5.0 ppm NO x , with corresponding levels of 5.0 ppm CO (corrected to 3% O 2 ), were achieved in a pulse combustor operating in a lean premixed mode, without the use of any postcombustion cleanup technologies. Both NO x and CO emissions were invariant to changes in the power input.