A test rig has been implemented for evaporator testing on a burner test setup. The setup comprises a water tank, fluid preheater, feed pump, large condenser, valves and all necessary instrumentation. Optionally an expander coupled to a dynamometer can be included for full Rankine Cycle testing as can be seen in

Fig. 3 CFD Flow uniformity validation for a 1,000 kWth evaporator unit with four parallel cores

Fig. 4 FEM design validation

Fig. 5. The test setup allows evaluation of evaporators, expanders and their inte- gration into complete systems.

4 Application

Bosal evaporators have seen in-field application since 2007. One early application integrated the evaporator for waste heat recovery in a railway power pack (Fig. 6). Other evaporators have accumulated operating hours in CHP generator sets for over 3 years now. An example of such an application is depicted in Fig. 7. In this application exhaust gas temperatures are on the order of 450 °C and water is used as working fluid. The generated steam is fed to a piston expander coupled to the crankshaft of a 15L V8 engine.

The evaporators have proven very durable, even though some engines were running on biofuel. For the CHP application the evaporator units are intermittently

Fig. 5 Rankine cycle test setup with expander and dynamometer

Fig. 6 Railway power pack with Rankine cycle waste heat recovery

Fig. 7 CHP generator set with evaporator mounted above engine

operated under high temperature dry conditions (absence of fluid flow) without sustaining damage.

5 Tools and Development

The near future will see more applications of Rankine Cycle Waste Heat Recovery. To offer suitable product proposals, a spreadsheet evaporator dimen- sioning tool has been developed and extended. The spreadsheet allows to find a suitable evaporator design configuration given an operating condition, desired heat exchange performance and exhaust back-pressure limitation. To meet the demand for alternative working fluids in Organic Rankine Cycle applications a calculation method has been devised based on the so-called effectiveness-NTU method. This approach based on non-dimensional numbers reveals key parameters: heat capacity ratio, dimensionless subcool temperature, Stanton heat transfer numbers and the Jakob number (sensitive to latent heat ratio). The method requires a minimal amount of fluid properties limited to: boiling point, vaporization heat, and specific heat capacity and transport properties at inlet, outlet and saturation con- ditions. It is also extended to allow predictions with working fluid mixtures.

A second software tool oriented development is the development of a dynamic evaporator simulation model. Current (Organic) Rankine Cycles are typically applied for stationary or mildly transient operation. Future systems, when applied on a heavy-duty truck for instance, will have to operate under transient conditions. The evaporator has its impact on the WHR system response that needs to be accounted for in system integration, optimisation and controls development [8, 9]. A transient simulation model of the evaporator is offered to support this system engineering task. The model is implemented in Simulink1 and can be integrated into other simulation environments like GT-SUITE 2 or AMESim3 .

Great care has been taken in the numerical implementation of the model. Simulation of two-phase flow and phase change phenomena can be challenging. Certain discretisation schemes and solution methods can lead to chattering; a phenomenon that causes oscillatory solutions [10]. It is intended to further develop a version of the simulation model that allows to study pulsive flow that can occur for certain setups. Insight into the nature of these highly dynamic phenomena will allow further improvement of the already durable design of the evaporator. For this purpose it is important that numerical simulation artefacts are not mistaken for real system dynamics.