Search

In this thesis, we describe our work with reverse automatic differentiation (AD) in the dataparallel programming language Futhark. The main motivation is to extend and optimise reverse AD in Futhark’s compiler to permit more expressive programs to by differentiated automatically and efficiently. Futhark’s AD consists of a set of rewrite rules that are used to transform a program to its differentiated counterpart. We present reverse mode AD rewrite rules for the operations reduce-byindex and scan. Reduce-by-index, also known as multi-reduce, has a generic cases and multiple special cases, of which the latter are loosely described by Schenck et al. 2022. We formulate and present a rewrite rule for the generic case and present specific rewrite rules for the special cases as Futhark pseudo-code. Likewise, we examine the reverse AD rewrite rules for scan presented by Schenck et al. 2022, one of which we have simplified with a performance benefit. The existing AD implementation is modified to work when the scan operates on tuples. We have extended the generic case with specialised rewrite rules for scan operators whose Jacobian matches specific patterns. We have implemented the presented rewrite rules of both reduce-by-index and scan in Futhark’s compiler. The performance of differentiated programs is evaluated experimentally and compared to its primal program performance and the program differentiated with forward AD instead. In many case, this demonstrates reasonable reverse AD overheads and competitive performance to Futhark’s established forward AD implementation.

When considering waste heat recovery systems for marine applications, which are estimated to be suitable to reduce the carbon dioxide emissions up to 20%, the use of organic Rankine cycle power systems has been proven to lead to higher savings compared to the traditional steam Rankine cycle. However, current methods to estimate the techno-economic feasibility of such a system are complex, computationally expensive and require significant specialized knowledge. This is the first article that presents a simplified method to carry out feasibility analyses for the implementation of organic Rankine cycle waste heat recovery units on board vessels using low-sulfur fuels. The method consists of a set of regression curves derived from a synthetic dataset obtained by evaluating the performance of organic Rankine cycle systems over a wide range of design and operating conditions. The accuracy of the proposed method is validated by comparing its estimations with the ones attained using thermodynamic models. The results of the validation procedure indicate that the proposed approach is capable of predicting the organic Rankine cycle annual energy production and levelized cost of electricity with an average accuracy within 4.5% and 2.5%, respectively. In addition, the results suggest that units optimized to minimize the levelized cost of electricity are designed for lower engine loads, compared to units optimized to maximize the overall energy production. The reliability and low computational time that characterize the proposed method, make it suitable to be used in the context of complex optimizations of the whole ship's machinery system.

As Arctic sea ice coverage declines it is expected that marine traffic could increase in this northern region due to shorter routes. Navigating in the Arctic offers opportunities and challenges for waste heat recovery systems (WHRS). Lower temperatures require larger heating power on board, hence a larger demand for waste heat usage, to cover services and maintaining on board spaces temperatures. However, a lower heat rejection temperature increases the WHRS thermal efficiency. The air temperature for the Arctic route selected is colder than that of the seawater, opening the opportunity of having air as coolant. This paper explores the use of two different coolants, air and seawater, for an organic Rankine cycle (ORC) unit using the available waste heat in the scavenge air system of a container ship navigating in Arctic Circle. Using a two-step single objective optimisation process, detailed models of air and seawater heat exchangers are evaluated as the WHRS condensers. The results suggest that an ORC unit using R1233zd(E) as its working fluid coupled with seawater as its coolant is the preferable option to reduce CO2 emissions. Using the ambient air as the coolant while a less effective option could be cheaper to install.

Electricity and power produced from waste heat is particularly relevant in shipping because fuel expenses constitute the majority of the cost of operating the ships; however, the cost-benefit aspect limits the widespread implementation of waste heat recovery power units on ships. This paper presents the thermodynamic analysis of a concept that aims at reducing the cost of an organic Rankine cycle unit by using one of the cylinders in a large diesel engine as expansion device. Numerical models were used to optimise the process parameters and thereby determine the power potential for this concept. The evaluation of 104 working fluids points to cyclopropane, R245fa and R1234ze(z) as the most promising. The results suggest that the power produced by the organic Rankine cycle cylinder is at least equivalent to that of the cylinders operating with the diesel process. This enables potential fuel savings and emissions reductions of about 8.3% in the studied scenario.

For zeotropic mixtures, the temperature varies during phase change, which is opposed to the isothermal phase change of pure fluids. The use of such mixtures as working fluids in organic Rankine cycle power plants enables a minimization of the mean temperature difference of the heat exchangers, which is beneficial for cycle performance. On the other hand, larger heat transfer surface areas are typically required for evaporation and condensation when zeotropic mixtures are used as working fluids. In order to assess the feasibility of using zeotropic mixtures, it is, therefore, important to consider the additional costs of the heat exchangers. In this study, we aim at evaluating the economic feasibility of zeotropic mixtures compared to pure fluids. We carry out a multi-objective optimization of the net power output and the component costs for organic Rankine cycle power plants using low-temperature heat at 90 ◦C to produce electrical power at around 500 kW. The primary outcomes of the study are Pareto fronts, illustrating the power/cost relations for R32, R134a and R32/R134a (0.65/0.35mole). The results indicate that R32/R134a is the best of these fluids, with 3.4 % higher net power than R32 at the same total cost of 1200 k$.

For zeotropic mixtures, the temperature varies during phase change, which is opposed to the isothermalphase change of pure fluids. The use of such mixtures as working fluids in organic Rankine cyclepower plants enables a minimization of the mean temperature difference of the heat exchangers whenthe minimum pinch point temperature difference is kept fixed. A low mean temperature differencemeans low heat transfer irreversibilities, which is beneficial for cycle performance, but it also results inlarger heat transfer surface areas. Moreover, the two-phase heat transfer coefficients for zeotropic mixturesare usually degraded compared to an ideal mixture heat transfer coefficient linearly interpolatedbetween the pure fluid values. This entails a need for larger and more expensive heat exchangers. Previousstudies primarily focus on the thermodynamic benefits of zeotropic mixtures by employing firstand second law analyses. In order to assess the feasibility of using zeotropic mixtures, it is, however,important to consider the additional costs of the heat exchangers. In this study, we aim at evaluatingthe economic feasibility of zeotropic mixtures compared to pure fluids. We carry out a multi-objectiveoptimization of the net power output and the component costs for organic Rankine cycle power plantsusing low-temperature heat at 90 C to produce electrical power at around 500 kW. The primary outcomesof the study are Pareto fronts, illustrating the power/cost relations for R32, R134a and R32/R134a(0.65/0.35mole). The results indicate that R32/134a is the best of these fluids, with 3.4%higher net powerthan R32 at the same total cost of 1200 k$.

The organic Rankine cycle is a suitable technology for utilizing low grade heat for electricity production. Compared to the traditional steam Rankine cycle, the organic Rankine cycle is beneficial, since it enables the choice of a working fluid which performs better than steam at low heat input temperatures and at lowpower outputs. Selecting the process layout of the organic Rankine cycle and the working fluid are two key design decisions which are critical for the thermodynamic and economic performance of the cycle. The prevailing approach used in the design and optimization of organic Rankine cycles is to model the heatexchangers by assuming a fixed minimum temperature difference. The objective of this work is to assess the applicability of this conventional optimization approach and a non-conventional optimization approach. In thenon-conventional optimization approach a total UA-value (the product of the overall heat transfer coefficient and the heat transfer area) is assigned to the cycle, while the distribution of this total UA-value to each of the heat exchangers is optimized. Optimizations are carried out for three different marine engine waste heatsources at temperatures ranging from 90 °C to 285 °C. The results suggest that the conventional optimization approach is not suitable for estimating the performance potential when the temperature profiles in the heat exchangers are closely matched. This is exemplified for the fluid MDM where the temperature profile of preheating aligns with the heat source fluid and for the zeotropic mixture R32/R134a where the temperature profile of condensation aligns with the cooling water. Furthermore, the conventional optimization approach shows weaknesses in evaluating the feasibility of using a recuperator, when the expander outlet temperature is high. In these cases the non-conventional optimization approach is the more suited methodology for designing organic Rankine cycles.

The international regulations on fuel efficiency and NOx emissions of commercial ships motivate the investigation of new system layouts, which can comply with the regulations. In combustion engines, measures to reduce the fuel consumption often lead to increased NOx emissions and careful consideration of this trade-off mechanism is required in the design of marine propulsion systems. This study investigates five different configurations of two-stroke diesel-based machinery systems for large ships and their influence on the mentioned trade-off. Numerical models of a low-speed two-stroke diesel engine, turbochargers and an ORC (organic Rankine cycle), are used for the optimisation of the NOx and fuel consumption at design and part-load conditions, using a multi-objective genetic algorithm. Moreover, the effects of engine tuning and exhaust gas recirculation are investigated. The results suggest that increased system complexity can lead to lower fuel consumption and NOx. Fuel consumption reductions of up to 9% with a 6.5% NOx reduction were achieved using a hybrid turbocharger and organic Rankine cycle waste heat recovery system.

Waste heat recovery systems can produce power from heat without using fuel or emitting CO2, therefore their implementation is becoming increasingly relevant. The Kalina cycle is proposed as an efficient process for this purpose. The main reason for its high efficiency is the non-isothermal phase change characteristics of the ammonia-water working fluid. The present study investigates a unique type of Kalina process called the Split-cycle, applied to the exhaust heat recovery from large marine engines. In the Split-cycle, the working fluid concentration can be changed during the evaporation process in order to improve the match between the heat source and working fluid temperatures. We present a system analysis to identify the governing mechanisms of the process, including a comparison of the efficiency of the Split-cycle and a conventional Kalina cycle and an investigation of the effects of using reheat in both cases. Results of a multi-variable optimisation effort using a genetic algorithm suggest that the Split-cycle process can obtain a thermal efficiency of 23.2% when using reheat compared to 20.8% for a conventional reference Kalina cycle. Reheat can increase the thermal efficiency by 3.4e5.9%. A simplified cost analysis suggests higher purchase costs as result of increased process complexity. © 2013 Elsevier Ltd. All rights reserved.

Much attention is focused on increasing the energy efficiency to decrease fuel costs and CO2 emissions throughout industrial sectors. The ORC (organic Rankine cycle) is a relatively simple but efficient process that can be used for this purpose by converting low and medium temperature waste heat to power. In this study we propose four linear regression models to predict the maximum obtainable thermal efficiency for simple and recuperated ORCs. A previously derived methodology is able to determine the maximum thermal efficiency among many combinations of fluids and processes, given the boundary conditions of the process. Hundreds of optimised cases with varied design parameters are used as observations in four multiple regression analyses. We analyse the model assumptions, prediction abilities and extrapolations, and compare the results with recent studies in the literature. The models are in agreement with the literature, and they present an opportunity for accurate prediction of the potential of an ORC to convert heat sources with temperatures from 80 to 360 C, without detailed knowledge or need for simulation of the process. © 2013 Elsevier Ltd. All rights reserved

Over the last years, much attention has been paid to the development of efficient and low-cost power systems for biomass-to-electricity conversion. This paper aims at investigating the design- and part-load performance of an innovative plant based on a wet indirectly fired gas turbine (WIFGT) fueled by woodchips and an organic Rankine cycle (ORC) turbogenerator. An exergy analysis is performed to identify the sources of inefficiencies, the optimal design variables, and the most suitable working fluid for the organic Rankine process. This step enables to parametrize the part-load model of the plant and to estimate its performance at different power outputs. The novel plant has a nominal power of 250 kW and a thermal efficiency of 43%. The major irreversibilities take place in the burner, recuperator, compressor and in the condenser. Toluene is the optimal working fluid for the organic Rankine engine. The part-load investigation indicates that the plant can operate at high efficiencies over a wide range of power outputs (50%–100%), with a peak thermal efficiency of 45% at around 80% load. While the ORC turbogenerator is responsible for the efficiency drop at low capacities, the off-design performance is governed by the efficiency characteristics of the compressor and turbine serving the gas turbine unit.

Strong motivation exists within the marine sector to reduce fuel expenses and to comply with ever stricter emission regulations. Heat recovery can address both of these issues. The ORC (organic Rankine cycle), the Kalina cycle and the steam Rankine cycle have received the majority of the focus in the literature. In the present work we compare these cycles in a combined cycle application with a large marine two-stroke diesel engine. We present an evaluation of the efficiency and the environmental impact, safety concerns and practical aspects of each of the cycles. A previously validated numerical engine model is combined with a turbocharger model and bottoming cycle models written in Matlab. Genetic algorithm optimisation results suggest that the Kalina cycle possess no significant advantages compared to the ORC or the steam cycle. While contributing to very high efficiencies, the organic working fluids possess high global warming potentials and hazard levels. It is concluded that the ORC has the greatest potential for increasing the fuel efficiency, and the combined cycle offers very high thermal efficiency. While being less efficient, the steam cycle has the advantages of being well proven, harmless to the environment as well as being less hazardous in comparison.

We present a generic methodology for organic Rankine cycle optimization, where the working fluid is included as an optimization parameter, in order to maximize the net power output of the cycle. The method is applied on two optimization cases with hot fluid inlet temperatures at 120°C and 90°C. Pure fluids and mixtures are compared to see how mixed working fluids affect performance and important design parameters. The results indicate that mixed working fluids can increase the net power output of the cycle, while reducing the pressure levels. The maximum net power output is obtained by fluids with a critical temperature close to half of the hot fluid inlet temperature. For some mixtures we find the maximum net power when the temperature glide of condensation matches the temperature increase of the cooling water, while for other mixtures there are large differences between these two parameters. Ethane is a fluid that obtains a large net power increase when used in mixtures. Compared to pure ethane, an optimized ethane/propane mixture attains a 12.9% net power increase when the hot fluid inlet temperature is 120_C and a 11.1% net power increase when the hot fluid inlet temperature is 90°C.

The Kalina split-cycle is a thermodynamic process for converting thermal energy into electrical power. It uses an ammonia–water mixture as a working fluid (like a conventional Kalina cycle) and has a varying ammonia concentration during the pre-heating and evaporation steps. This second feature results in an improved match between the heat source and working fluid temperature profiles, decreasing the entropy generation in the heat recovery system. The present work compares the thermodynamic performance of this power cycle with the conventional Kalina process, and investigates the impact of varying boundary conditions by conducting an exergy analysis. The design parameters of each configuration were determined by performing a multi-variable optimisation. The results indicate that the Kalina split-cycle with reheat presents an exergetic efficiency by 2.8% points higher than a reference Kalina cycle with reheat, and by 4.3% points without reheat. The cycle efficiency varies by 14% points for a variation of the exhaust gas temperature of 100 °C, and by 1% point for a cold water temperature variation of 30 °C. This analysis also pinpoints the large irreversibilities in the low-pressure turbine and condenser, and indicates a reduction of the exergy destruction by about 23% in the heat recovery system compared to the baseline cycle.

The Synthetic Aperture Real-time Ultrasound System (SARUS) for acquiring and processing synthetic aperture (SA) data for research purposes is described. The specifications and design of the system are detailed, along with its performance for SA, nonlinear, and 3-D flow estimation imaging. SARUS acquires individual channel data simultaneously for up to 1024 transducer elements for a couple of heart beats, and is capable of transmitting any kind of excitation. The 64 boards in the system house 16 transmit and 16 receive channels each, where sampled channel data can be stored in 2 GB of RAM and processed using five field-programmable gate arrays (FPGAs). The fully parametric focusing unit calculates delays and apodization values in real time in 3-D space and can produce 350 million complex samples per channel per second for full non-recursive synthetic aperture B-mode imaging at roughly 30 high-resolution images/s. Both RF element data and beamformed data can be stored in the system for later storage and processing. The stored data can be transferred in parallel using the system’s sixty-four 1-Gbit Ethernet interfaces at a theoretical rate of 3.2 GB/s to a 144-core Linux cluster.