Sustainable power generation with large gas engines

Energy Conversion and Management xxx (2017) xxx–xxx

Contents lists available at ScienceDirect

Energy Conversion and Management journal homepage: www.elsevier.com/locate/enconman

Sustainable power generation with large gas engines Gerhard Pirker a,⇑, Andreas Wimmer a,b a b

LEC GmbH, Inffeldgasse 19, 8010 Graz, Austria Institute of Internal Combustion Engines and Thermodynamics, Graz University of Technology, Inffeldgasse 19, Graz, Austria

a r t i c l e

i n f o

Article history: Available online xxxx Keywords: Gas engine Power generation Climate change mitigation Alternative fuels Gas quality Grid gas Development methodology Combustion system

a b s t r a c t Large gas engines will play a significant role in distributed power generation for the energy supply of the future. The lower amount of carbon in natural gas in comparison with other fossil fuels can be used to bridge the gap between a carbon-based and a carbon-free energy supply. The main objective of this paper is to provide an overview of the technological challenges the next generation of gas engines will face. Improvements in robustness and dynamic behavior will allow gas engines to meet the high transient requirements for the future power supply. The great fluctuations in gas quality anticipated with grid gas and liquefied natural gas impose high demands on both the transient behavior and the knock resistance of the engine. Technologies that enhance fuel flexibility by enabling sustainable power and heat generation using hydrogen-rich syngas from biomass and the efficient use of waste gases will be key. The most important technological components that maximize power output and efficiency as well as transient operation at very low emission levels are discussed. An advanced development methodology is applied in order to deal with the requirements presented by the technological challenges. The main future goals of gas engine development will be described by use of examples to illustrate the application of the methodology. In summary, the research and technological developments presented in this paper will support the transition from conventional to carbon-free fuel for reliable and sustainable power generation that meets future requirements for large gas engines. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction In the power generation systems of the future, the fraction of renewable energy sources such as solar power and wind power will increase. The dependence of these power sources on the weather as well as the decentralization of the energy market are increasing demand for fast-reacting and well-controlled sources of electrical energy to compensate for peak loads or drops in power generation. In this scenario, the difference between production and consumption of electrical energy will continue to increase. In general, low-carbon energy production strategies have to be pursued. One example of such a strategy can be found in a recent study [1] comparing the perspectives for low-carbon energy production in two EU countries in the next few decades. Different strategies, i.e. renewable energy sources (wind, hydro, solar, bioenergy), carbon capture and sequestration strategies, are assessed and recommendations based on the regional boundary conditions are made with consideration of the cost of electricity. ⇑ Corresponding author. E-mail addresses: [email protected] (G. Pirker), [email protected] (A. Wimmer).

Internal combustion engines will play a significant role in decentralized power generation for the energy supply of the future because they are able to quickly react to fluctuations in demand for power and incorporate a compensation measure to stabilize the electric grid. From an environmental point of view, gas engines are a key factor in power generation due to their comparatively low environmental impact, which is further enhanced by the great progress in terms of efficiency and power output that has been achieved in recent years. A number of studies show that the global share of natural gas for power generation will continue to develop, cf. [2], which analyzes the challenges emerging from the fuels of the future for internal combustion engines and assesses feasible fuel options. A prognosis made by the International Energy Agency (IEA) about the development of global power generation as divided into shares of the individual energy carriers (World Energy Outlook 2012) [2] shows that the share of natural gas in power generation will increase significantly in the next 20 years. This trend is mainly driven by the large amount of natural gas reserves in conventional and unconventional deposits (e.g., shale gas, tight gas and coal bed methane). The discussion about a strategy to phase out nuclear energy in various countries (e.g., Germany and Japan) as well as the

http://dx.doi.org/10.1016/j.enconman.2017.06.023 0196-8904/Ó 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

2

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

trend towards lower natural gas prices are also encouraging the use of gas. Moreover, natural gas helps to bridge the gap between carbon-based and carbon-free energy sources. Various applications for gas engines currently exist. One of the most important gas engine applications is the combined heat and power plant, where combustion engines produce electricity and heat. Many power generation applications of gas engines are in use that are independent of the grid, in particular generator sets and mechanical drive applications for pumps and compressors (e.g., petrochemistry, oil and gas production, wastewater treatment). In addition, special gases such as landfill gas, waste gases from industry, and flare gas are more frequently exploited as sources of fuel for gas engines. Recent studies [3,4] evaluate how these special gases can be applied in internal combustion engines. 2. Sustainability Although gaseous fuels for power generation are already widely available today, numerous signs indicate that attitudes towards their use continue to change around the world. Sustainable power generation from renewable sources instead of classic fossil fuels or nuclear fission is a global goal. The EU Climate & Energy Framework 2030 has set three targets: that greenhouse gas emissions fall to levels at least 40% lower than the 1990 levels, that the share of renewable energy increases to 27% and that energy efficiency increases by at least 27%. To reach these targets, the power sector in particular will need to reduce greenhouse gas (GHG) emissions drastically [5]. To be sustainable, it is imperative that electrical power is generated from a combustion process that is either CO2 neutral or produces as little CO2 as possible. In addition, overall pollutant emissions should be as low as possible. In the area of power and heat generation from combustion engines, one promising path towards achieving sustainability is to make the transition from carbon-based to carbon-free energy sources. 2.1. Transition to carbon-free fuels At the start of the industrial revolution, ‘‘solid” sources of energy such as coal were most commonly used to power industry. Mainly driven by the demands of mobility, the advantages of liquid fuels became obvious and their advantages were soon exploited for power generation as well. At present, natural gas offers further advantages; not only are carbon dioxide emissions lower than with liquid fuels but they also have the potential to produce even fewer pollutant emissions [2]. Complete combustion of hydrocarbons always produces carbon dioxide and water. As oxidation of hydrogen yields water and oxidation of carbon yields carbon dioxide, the hydrogen-to-carbon ratio of the fuel determines the fraction of carbon dioxide in the exhaust gas. As can be seen in Fig. 1, the hydrogen-to-carbon atomic ratio of coal is approximately 0:1 while this ratio for gasoline or diesel is roughly 2:1. The hydrogen-to-carbon ratio of natural gas, whose main component is methane (CH4), is about 4:1. Hence, the combustion of natural gas produces less CO2 per unit of energy than either coal or gasoline [2]. The transition from solid coal to liquid hydrocarbons and from natural gas to pure hydrogen as a carbon-free fuel can therefore be regarded as a decarbonization pathway [2] and must be pursued further. In this context, due to its CO2 neutral character, synthetic methane can also be regarded as carbon-free in consideration of the production and consumption cycle. However, the production of hydrogen may still require consumption of fossil fuel sources such as oil and gas when current technologies are applied. Therefore, it is imperative to produce hydrogen using renewable energy sources such as wind power or solar power.

2.2. Power generation with large gas engines A key target in developing large gas engines for power generation is to obtain the highest possible efficiency. Recent studies [6,7] describe the measures undertaken to attain a remarkable increase in the efficiency of high performance gas engines for distributed power generation. In addition to the increase in efficiency and power of gas engines acquired in recent years, the smaller carbon fraction of natural gas in comparison with other fossil fuels is another factor that enables gas engines to be applied to bridge the gap between carbon-based and carbon-free energy sources. Gaseous fuels for engine combustion can be divided into fossil gases, biologically produced gases and technically produced gases. The gases are classified according to their origin, which influences their compositions and combustion properties. In this context, three main paths to achieving sustainability emerge. All three paths have a different impact on engine combustion, present different challenges to the engine and therefore require different engine technologies. Fig. 2 provides an overview of these paths and their impact on engine operation. Three main paths with several variations can be distinguished: 1. The ‘‘classic” path still uses fossil fuel sources but seeks to optimize the use of these fuels. Fuels with a high hydrogen to carbon ratio such as natural gas are a good choice since their CO2 emissions are lower than those of liquid fuels. Furthermore, measures must be taken that increase the efficiency of the engine as well as reduce pollutant emissions of all kinds. The main challenge of this path is to continue to develop combustion systems that achieve the highest possible efficiencies with the lowest possible emissions. 2. A second path is to use fuels that would otherwise not be exploited, e.g., waste gases from industrial processes. One good example is blast furnace gas (BFG), which is created during the steel production process. Flare gas that leaks from petroleum refineries, natural gas processing plants and oil or gas production sites is another type of waste gas. Instead of it being burned off in gas flares, its energy content can be used for power generation and to operate the plant itself. Since flare gas widely varies in its composition depending on where and how it is produced and how much is produced, optimal design of gas engines is critical. 3. The third and most sustainable path is to use renewable fuels. For example, hydrogen can be produced using renewable sources such as wind power or solar power. This hydrogen can be exploited directly or converted into synthetic methane using carbon dioxide from industrial processes by applying the methanation process. This synthetic methane can then be used instead of natural gas. Biomass can be converted into biogas by gasification processes as well as anaerobic digestion. Thus, the composition of biogas greatly varies depending on its source. All these renewable gases can either be added to the gas grid or directly fed to the engine. Gaseous fuels from renewable sources present different challenges for the engine depending on how they are produced and used. 3. Challenges Along the road to sustainable power generation, technological challenges have to be overcome and the requirements for the engine have to be met by research and application of highly sophisticated engine technologies. Improvements in robustness and dynamic behavior will allow gas engines to compete with diesel engines in applications with high transient requirements caused by fluctuations in the electric grid. The great fluctuations in gas quality anticipated with grid gas impose high demands on both

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

3

Fig. 1. Energy mix transition.

Fig. 2. Sustainable power generation.

transient behavior and knock resistance of the engine. A key factor in the sustainable energy supply of the future will be the operation of gas engines with a variety of renewable gases or gases that would otherwise not be exploited. 3.1. Emission limits All future research and development efforts in the area of combustion engines will be heavily influenced by developments in emission limits. Emission regulations for large engines differ greatly around the globe; limits are defined depending on the region and the application. As NOx emissions arise also from combustion of carbon-free fuels, the permissible limits are also an issue for engines that burn alternative fuels. Moreover, the level of NOx emissions has a great influence on achievable efficiency and engine technology. Fig. 3 provides an example of the future development of NOx limits in the EU for stationary lean burn gas engines, diesel engines and dual fuel engines (i.e. engines that can burn natural gas and diesel). It shows the development of limits in the Gothenburg Protocol [8] and EU directive 2010/75/EU [9] in comparison to TA Luft [10]. TA Luft is an air pollution control regulation (Technische Anleitung zur Reinhaltung der Luft) originally established in

Germany that sets emission limits for various air pollutants. It has been updated many times over the past 50 years and adopted by other countries. With the exception of a few regions, the EU limit of 200 mg/m3n (with a 5% O2 concentration in the exhaust gas) is the strictest limit for lean burn gas engines. This limit is the basic goal for stationary applications, yet it is imperative to achieve it without losses in efficiency. As a result of the recent discussion of emissions of formaldehyde, which is suspected to be a carcinogen, gas engines face yet another series of challenges. According to TA Luft [10], emissions of formaldehyde are limited to 60 mg/m3n; if formaldehyde is proven to be a carcinogen, permissible emissions should be limited to 1 mg/m3n, more details can be found also in [12]. Gas engines with homogeneous charge mixing or port injection and a premixed combustion generally emit significantly higher amounts of formaldehyde than diesel engines or engines with a diesel-like diffusion based combustion. A detailed discussion of formaldehyde emissions of combustion engines can be found in the CIMAC position paper [13]. Formaldehyde emissions are especially an issue with engines burning biogas, which can only be used after the gas has undergone an extensive purification procedure to remove

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

4

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

Fig. 3. Various NOx emission limits for stationary gas engines [11].

catalyst poisons. In the future, it will become increasingly important to avoid the formation of this pollutant within the engine. In recent years, the topic of the environmental impact of methane slip has also been intensely discussed, cf. the position paper of the European Association of Internal Combustion Engine Manufacturers on methane slip from internal combustion gas engines [14]. Methane as the main component of natural gas makes up the largest amount of hydrocarbon (HC) emissions and its impact on the greenhouse effect is around 25 times greater than that of CO2. From the standpoint of total greenhouse gas emissions as well as efficiency, it is essential to reduce the level of HC emissions from the engine to a minimum; this reduction is thus an important research objective.

3.2. Increased output and efficiency Sustainability requires that available resources must be used as efficiently as possible. The efficiency of gas engines has rapidly developed to catch up with that of diesel engines and even exceeds it especially at low NOx limits. Since the efficiency of gas engines is already high, any further increase requires a large amount of effort and presents a great challenge in combustion concept development. From a thermodynamic point of view, it is possible to improve the efficiency of gas engines by increasing the compression ratio even more. High air/fuel ratios and Miller valve timing place great demands on the turbocharger. However, thermodynamic measures will not be sufficient to achieve a further increase in performance and efficiency. The turbocharger has become a key component for obtaining high engine efficiency. This is also the main conclusion of many contributions that deal with further measures to increase the efficiency of large gas engines. The investigations in [15] describe the current development status of high-speed and medium-speed large gas and diesel engines in terms of efficiency and analyze the trends in the development of efficiency and measures to comply with future emission legislation. The article comes to the conclusion that due to future emission regulations the efficiency of diesel engines can hardly be further improved by conventional methods. In the field of gas engines only the lean burn concept seems to be a promising way for further increase in efficiency despite its drawbacks in NOx emissions. The turbocharger will again have a large impact on

these improvements. Another publication [16] describes the potential of two-stage turbocharging as a means to increase engine efficiency. Again, the turbocharging system is identified as the major contributor to achieve increased power density, reduced emissions and high total engine efficiency. For optimal engine performance the authors of this contribution recommend a high pressure single stage turbocharger or a two-stage turbocharging system. For a further increase of efficiency a variable valve train system is required, which does not only allow the optimization of relevant operating points, but also enables an optimization of the whole engine map which is mandatory for transient engine operation. The issue of transient requirements will be thoroughly discussed later in this paper. The authors of [17] describe the possibilities of an electro-hydraulic variable valve timing (VVT) system and present the thermodynamic performance acquired by de-throttling. Furthermore the advantages of the VVT system if applied to engine operation in normal conditions as well as during load changes are discussed. Besides the turbocharging and the VVT systems also wear and durability issues will greatly influence power output and efficiency, thereby requiring an advanced development methodology such as the one explained later in this paper. By applying a combination of new technologies and optimization measures the efficiency of a modern gas engine in the 10 MW power range has already exceeded 50%. A comprehensive contribution [18] outlines the required technology blocks including combustion concept, air path, power unit, turbocharging module, cylinder head and controls, and concludes that the operation of the engine has fully entered the digital age with remote connectivity, diagnostics and advanced data analytics to provide lowest maintenance cost and best availability.

3.3. Requirements for transient engine operation Combustion engines are a good choice to compensate for peak loads or drops in power generation caused by the dependence of renewable energy sources on the weather. Load changes have to be made very quickly in order to stabilize the grid. Diesel engines manage a load increase from 0% to 100% in less than 60 s depending on engine size and power output. The values for gas engines are much higher and exceed 100 seconds for very large gas engines.

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

20 s and two to three load steps can be defined as target values for gas engines in the up to 5 MW power class. For larger engines, 40 s seem to be realistic. The implication is that significant improvements in transient behavior are necessary if gas engines are to be used in stationary operation (island mode, grid parallel operation) given the increasingly strict future requirements. Since the load acceptance of the gas engine is greatly limited by the knock limit, the dynamic behavior of the gas engine is inferior to that of a conventional diesel engine. The challenge of providing load acceptance is even greater with high performance gas engines, which achieve a high level of efficiency and BMEP and operate with a very high compression ratio, extreme Miller valve timing and a high level of boost pressure. An overview of the demands in terms of transient response behavior of gas engines can be found in the position paper [19] of the CIMAC (Conseil International des Machines A Combustion) Working Group. Furthermore, several technical standards define the requirements for transient operation of combustion engine operating as gensets in grid parallel mode as well as in island mode. The ISO 8528-5 technical regulation [20] generally specifies design and performance criteria arising out of the combination of combustion engine and an alternating current generator when operating as a unit (genset). However, there are several exceptions for special applications, e.g., essential hospital supplies [21] and fire protection [22]. All these requirements impose additional challenges to the engine that have to be overcome by advanced technologies. In general, transient engine operation can be divided into several phases, i.e. engine start, steady state operation, load increase, load decrease, grid faults and engine stop. For each phase, additional requirements exist. For engine start, the contribution of [23] states that today’s benchmark time for just loading is five minutes, but it is foreseen that this goes down to two minutes in the very near future. For black start and island operation it shall be possible to apply 10% load steps with a frequency drop not exceeding 1%. Black start means that the power plant is not connected to any grid (cold or heated up engine) at startup. The authors of [24] state an acceleration capability at a minimum of 2% power increase per second, which results in the time from idling (synchronised to the mains) to full load in no more than 50 seconds. The network code for requirements of grid connection [25] is applicable to all power generators and defines a common framework of grid connection requirements for power generating facilities. A maximum admissible initial delay of two seconds and a full activation time of 30 seconds until synchronization to the grid are prescribed. For steady state operation as well as load increase and decrease, several regulations exist. According to [20] the engine has to stay within a frequency tolerance band at load increase and decrease. In steady state operation, a stable genset power output is required. The fluctuations should not exceed 1% over 1 min or 30 min [23]. The authors of [24] even define a steady state frequency fluctuation margin of 0.2% (±2 rpm for 1000 rpm) for special applications. In order to comply with the requirements imposed by grid faults, the genset must not disconnect from the grid if a drop in grid voltage appears for a short period of time. For this application case there exist several regional regulations. In France, the engine has to withstand one voltage drop within a period of 150 milliseconds [26]. According to a technical regulation [27] in Germany, the engine has to withstand two voltage drops, where one voltage drop can last 150 milliseconds. After five seconds, the engine has to recover and deliver at least 95% of the full power to the grid again. Regarding engine stop, load rejection from full load to idle has to be accomplished as fast as possible in order to be able to reconnect to the grid at any time, which is stated in both contributions [23,24].

5

To sum up, the vast diversity of transient engine operation regulations require advanced technologies to improve dynamic behavior of a gas engine which will be discussed later in this paper. 3.4. Robustness Along with the worse dynamic behavior compared to diesel engines, the gas engine also has the disadvantage that engine behavior changes over runtime. This change is mainly caused by wear on the components involved in combustion and the formation of deposits on the combustion chamber walls. These effects can even be worse when alternative fuels such as biogas or hydrogen are used. The limited service life of the spark plug shortens the maintenance interval for gas engines and represents a further drawback. Since it is critical that combustion concepts are robust, their design should be based on a profound understanding of wear mechanisms so that components can be optimized according to the higher demands placed on high performance engines. To this end, it is of prime importance to design robust combustion concepts. Essential to this design is the development of a profound understanding of wear mechanisms so that components can be optimized purposefully. The continuously increasing demands being placed on the robustness of highly stressed engine components necessitate an improvement in simulation methods. Simulation is already standard in component optimization since it can explain effects and allows a better overall understanding. This is one of the conclusions of the publication [28] that gives a neat overview of the possibilities and limitations of simulation methods in the field of component optimization. The authors further conclude that coupled CFD (computational fluid dynamics) and FE (finite element) calculations impose further challenges to the requirements of accurate boundary conditions. Anyway, coupled simulation methods can be regarded as powerful tool. A further contribution [29] also states that in many cases simulation results are influenced by uncertainties in the boundary conditions applied. The applied simulation models and material properties have to be selected properly. 3.5. Renewable feedstock and integration into an electric grid or gas grid The challenges that arise from using renewable power sources for combustion engines are highly diverse. Depending on the power generation path, cf. Fig. 2, the electrical power generated from wind energy or solar energy can be either fed directly into the electric grid or converted to gaseous fuels such as hydrogen or methane. If electrical power from solar energy and wind energy is fed directly into the electric grid, combustion engines are often chosen to compensate for peak loads or drops in power generation caused by the dependence of renewable energy sources on the weather. In this case, it is challenging to meet the requirements for combustion concepts, control concepts and load management concepts for highly transient engine operation, cf. section 3.3. On the other hand, it is possible to store electrical power in chemical form by converting the excess power or off-peak power generated by wind turbines or solar arrays to hydrogen so that this power may be exploited at a later time. The gas grid can be used to store and transport the hydrogen. As a consequence of hydrogen being fed into the gas grid, the combustion properties of the grid gas are no longer constant and may change significantly over time. These fluctuations in gas quality present a great challenge to combustion engines. Alternatively, the hydrogen from the electrolyzer can be combined with CO2 (e.g., from industrial processes) and converted to methane using a methanation reaction. This synthetic methane

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

6

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

can then be fed into the gas grid with only small changes in the quality of the gas in the grid. Another challenge in the sustainable application of gas engines is how to use synthetic gas generated from the conversion of biomass and biodegradable waste. Since current synthesis processes are already very sophisticated, this gas can contain a large fraction of hydrogen and a very small fraction of pollutants. Gas compositions with high hydrogen content require quite low ignition energies and provide high combustion velocities. This leads to increased probability of knocking and undesirable backfiring behavior, thereby limiting economical use of these gases. The challenge is to develop a gas engine combustion concept that can deal with such demanding gas compositions while allowing economical and CO2 neutral power generation. This challenge is also indicated in the studies [4,30] that deal with the use of special gases from organic material in stationary gas engines. Possible target values for such a concept include achievable power output, efficiency, emission limit as well as sufficient combustion stability and robustness in terms of knocking and backfiring. To sum up, if gases from renewable sources, i.e. either from power-to-gas processes or biomass conversion, are directly fed into the engine, certain measures have to be taken to design a combustion concept that can make use of gases with an H2 fraction of more than 40%. If hydrogen or hydrogen rich gases are fed into the gas grid, fluctuations in gas quality occur which in turn present challenges to all engines fed by gas in the grid. In order to overcome the impact of fluctuating gas qualities on current and future gas engine technologies, regulations for gas quality are desired. There exist many different standards and regulations in the EU and the US which impose additional challenges to the engine manufacturer. In the EU each country has its own regulation in the field of natural gas transport within a gas pipeline network or gas distribution system and the adding of biogas. However, there is a new European standard for the quality of gas released from CEN (European Committee for Standardization). In the US, the regulation of the natural gas transport is handled through the FERC (Federal Energy Regulatory Commission) and the pipeline and distribution companies. Regulations should make reference to either the calorific value or the Wobbe Index, each of which must be maintained within a narrow range. However, for the latest European standard EN16726 [31] no agreement regarding the Wobbe Index could be reached. So the European Commission releases the standard without limits for the Wobbe Index. Nevertheless the planned limits for the Wobbe Index, which were shown in the draft of the regulation, are included. Further information on the limits of CEN [31] and EASEE (European Association for the Streamlining of Energy Exchange) [32] can be found in the literature. The change in methane number, which has a strong impact on the knock behavior of gas engines, is not yet regulated. At present, a lower limit for the methane number is under discussion in a paper on gas quality harmonisation by the European Association of Internal Combustion Engine Manufacturers [33]. The knock behavior is critical to gas engine operation, therefore several criteria for the assessment of knock behavior will be discussed later in this paper. 3.6. Waste gases from industrial processes Large gas engines already put non-natural gases (e.g., biogas, steel gas, BFG, landfill gas, flare gas, mine gas) to use for combined heat and power generation. This approach permits an environmentally sound and energy efficient use of gases which would otherwise not be exploited. Examples of gas engine applications that in particular use waste gases can be found in [34]. An engine series capable of burning non-natural gases is presented in the contributions [35,36]. In these contributions the engine concept, the design

features, and the special adaptations of the engine for use of nonnatural gas are explained. The authors conclude that the new engine series succeeded in extending the range of power ratings for biogas engines to 2 MW as well as the design of the engines offers further potential for performance and efficiency gains. In refinery operations, flammable waste gases are vented from processing units during process upsets and normal operation. These waste gases are collected in piping headers and delivered to a flare system for safe disposal. Studies show that approximately 150 billion cubic meters of gas are flared globally each year, wasting natural resources and generating 400 million metric tons of CO2 equivalent global greenhouse gas emissions [37]. Since flare gas widely varies in its composition depending on where and how it is produced and how much is produced, optimal design of gas engines is critical. Creating more flexibility for these stationary applications will make more efficient and appropriate installations possible. The special properties of flare gases present a major challenge in developing efficient and robust combustion concepts for large gas engines. Furthermore, gases with a very low calorific value (LCV) are interesting as a future source of engine fuel. When steel is produced using the blast furnace process, very large amounts of waste gases are generated. Coke gas that forms in the coke oven battery and converter gas released when the carbon content of pig iron is reduced by oxygen can be successfully exploited for power and heat generation because of their relatively good combustion properties. In contrast, the use of blast furnace gas (BFG) is proving to be more difficult because of its extremely unfavorable properties [3,38]. In this context, dual fuel technology could have advantages. Nowadays the development status of gas engine applications using mine gas, biogas, landfill gas and sewage gas has reached a level comparable to natural gas applications. If the engines are properly calibrated to cope with the special requirements of the gas, power output and efficiency can reach levels almost identical to those of natural gas applications without requiring any further effort. The progress that has been made in the application of gases that contain a large share of hydrogen and CO in high efficiency gas engines is presented in [39]. However, engine operation with steel gas or synthesis gas is a greater challenge. These engines have to be thoroughly adapted to the special characteristics of the gas, which also involves modifying the combustion system and engine control. Power output must often be reduced to about 60% of the power output with natural gas operation. The challenges of using waste gases are mostly linked to the very strongly fluctuating gas compositions and the inconstant availability of these gases. As outlined above, the varying gas properties, e.g., lower calorific value (LCV) and methane number, require very robust combustion concepts and sophisticated control strategies.

4. Required technologies 4.1. State-of-the-art performance of large engines To meet the challenges discussed in the previous section, innovative technologies that fulfill the requirements for large gas engines must be investigated. Evaluation of the technologies currently available aids in identifying potential areas of gas and dual fuel engines that can be developed in this direction. Fig. 4 shows selected features in the areas of performance and robustness. Performance is characterized by load response, power output, emissions and efficiency while robustness consists of long-term stability, fuel flexibility and maintenance. These challenges and demands are compared in [40]. To determine the position of gas engines among other combustion engines, the characteristics of

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

7

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

Performance Efficiency

Diesel Engine Gas Engine Dual Fuel Engine

Gas Mode Diesel Mode

o + o --

Robustness

Emissions NOx

PM

CO2

CH4

Power density

+ + -

-++ o --

+ + --

++ ++

++ + + +

Long term Load Maintenance stability response

++ o ++

++ + +

+ o

Fuel flexibility

+ +

Fig. 4. Development status of engine technologies [40]

diesel engines are also provided in the figure. From the point of view of sustainability, the gas engine has fewer emissions of all kinds except CH4 emissions due to methane slip in the scavenging phase and the loss of unburned gas during the exhaust phase. In recent years, the gas engine has outperformed the diesel engine in terms of efficiency. However, it still has significant disadvantages in the areas of load response and robustness. Thus efforts to investigate new gas engine technologies to lower emissions should mainly focus on reducing methane slip. Nevertheless, further innovative engine technologies aimed at improved efficiency and low emissions should still be pursued, e.g., variable valve timing, turbocharging, gas direct injection into the pre-chamber and ignition systems. Even greater efforts have to be invested into both a new comprehensive approach that meets the demands of transient operation and robust combustion systems that are able to deal with various gas compositions as well as fluctuations in the quality of grid gas.

4.2. Reduction of methane slip Not only CO2 emissions but also unburned hydrocarbons significantly contribute to the greenhouse effect and reduce engine efficiency. As stated above, highly efficient gas engines with low NOx emissions may emit increased HC emissions, which consist mainly of unburned methane. Although exhaust gas aftertreatment of methane emissions is theoretically possible, it is difficult to apply due to the chemical stability of methane. Furthermore, any unburned fuel in the exhaust worsens the efficiency of the engine. Therefore, it is imperative to reduce unburned HC emissions inside the engine by optimizing the combustion system. Research into this technology has to identify the primary sources of HC emissions from gas engines and define measures to avoid them. For a gas engine that complies with the current NOx emission standards, a number of measures can be taken to reduce HC emissions and achieve high efficiencies. The authors of [41] thoroughly study the formation of HC emissions, presenting simulation methods and reduction measures. Further simulation methods that support the development of strategies for the HC reduction in the exhaust gas can be found in [42]. The authors present a zonal cylinder model for gas engines fueled by methane. The model simulates the combustion and attains results like the heat release rate, the cylinder pressure and the temperatures within the combustion chamber. Further results are the quenching distances of the flames which are critical to unburned HC emissions. By applying their model they conclude that for the methane gas engine the quenching distances are in the range of 23–61 lm. The quenching distance highly depends on the pressure, not on the wall temperature. These investigations imply that combustion chamber geometry has to be designed in a way that avoids crevices.

Furthermore valve timing must be optimized to prevent the scavenging of fresh gas into the exhaust. Charge air cooling, Miller valve timing and optimized turbocharging efficiency can help to increase the specific load and lead to high mean effective pressures. Optimized mixture preparation and an improved ignition system and ignition timing should guarantee stable ignition of the charge. In addition, a suitable prechamber supports reliable ignition of the mixture. Increasing turbulent combustion velocity helps with lean operating points or operating points that burn biogas with a low LCV in particular. Finally, injection of the fuel gas directly into the combustion chamber has the potential to greatly reduce methane slip.

4.3. Improvement of dynamic behavior Improved transient behavior of the engine is mandatory to comply with the challenges of power fluctuations or gas quality fluctuations imposed by the integration of renewable feedstock into an electric grid or gas grid. This also comprises strategies for reducing the emissions of the engine in transient operation. Improved dynamic behavior could be achieved through many measures like reducing the load, changing the air/fuel ratio, reducing the compression ratio or increasing the control reserve. These measures would result in a more robust engine design but poorer efficiency and emission values; therefore, this strategy is not desirable. One major goal is to design high performance gas engines with sufficiently transient behavior. To improve dynamic behavior, expanded adjustable parameters in the air path as well as an optimized control strategy are required. When it comes to dynamic performance, again the turbocharger becomes an important device. Therefore measures to support the turbocharger acceleration are essential, which are discussed in [43]. The authors present technologies that comply with these demands and conclude that mainly two technologies, i.e. a Jet Assist system and an electrical boost, show promising results with the positive side effect that the efficiency of the engine in steady-state operation does not suffer. A further contribution [44] presents a concept study how the two stage turbocharging system could be designed in the near future. The authors compare various technology blocks, evaluate them, and conclude that from the transient performance point of view variable valve train and the integrated turbo compounding solutions would deliver best results. In [45] also the most important technology requirements are listed. Advanced intake valve closing, high mixture temperature and high compression ratios as well as low pressure gas injection, passive prechamber system and highly efficient gas exchange and turbocharging system have to be considered. The contribution concludes that a detailed system analysis, thorough simulation activities, and hardware upgrades including control concepts enable the stability of

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

8

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

the combustion process throughout the complete transient event. An example of the development of a large engine that complies with transient requirements can be found in [46], where all required technology blocks are explained. The authors conclude that the developed engine is well suited for the production of electricity and heat on demand with a response time to full load in less than five minutes being the main entry criteria to serve in ancillary segments. Further, the engine is capable to operate as the single power generation unit in isolated grids ensuring high stability and low impact on grid frequency and voltage in case of fast load steps. For reasons of time and cost, it is imperative to design and evaluate combustion, valve timing and control concepts for transient operation of large engines early in the development process, preferably by simulation or at a single cylinder engine. General approaches can be found in the literature for how to apply the ‘‘ Hardware-in-the-loop” (HiL) method for testing and evaluating of embedded systems during engine development [47]. The authors describe the general considerations of real-time simulators and architectures of HiL simulation, present an application example of the HiL development of a heavy-duty engine and conclude that the methodology can be used as an efficient tool to develop new control functions, to test the software and hardware of the engine electronics. More recent publications, cf. e.g., [48,49], describe transient test systems of SCE for automotive applications. In [48] the focus lies mainly on the design of the hardware (dynamometer, intake airflow simulator) for passenger car sized engines and the contribution concludes that these systems can be applied to make an SCE to operate like it were part of a multicylinder engine. In [49] the same equipment is further developed; the focus is on the intake airflow simulator development, including other applications in simulating turbocharged or supercharged engines. Possibilities, limitations and control strategies are outlined. The applied models are mainly based on look-up tables. A recent publication [50] provides an overview of the application of HiL models on heavy duty diesel engines. A more extensive discussion in [51] shows the most complete approach for the development of transient engine behavior, the authors also conclude that the complexity of all involved systems require a development methodology based on a HiL setup. Nevertheless, the application of HiL models on heavy duty diesel engines is shown, the authors do not consider the special requirements for large gas engines. The authors of [52] start with a detailed model in an engine cycle simulation program and subsequently simplify the components in order to achieve real-time capability with a so-called fast running model for investigating MCE behavior in automotive applications. They conclude that a detailed simulation model can be simplified into a fast running model that runs in real-time on a HiL platform. In contrast to the efforts in the area of passenger cars and heavy-duty applications, the authors of [53] describe the development of a methodology to simulate transient MCE behavior of large gas engines on a SCE test bed. Simple and fast models and algorithms are created that are able to provide the boundary conditions (e.g., boost pressure and exhaust back pressure) of a MCE in transient operation in real-time for use on the SCE test bed. To sum up, technologies to improve the transient behavior of an engine lie mainly in the improvement of variable valve timing, advanced turbocharger layout and control strategies, as well as transient combustion concepts that comply with future emission regulations and efficiency requirements. More details as well as an example of the HiL operation of an SCE to optimize the transient behavior of large gas engines combined with an advanced development methodology can be found in Section 5.2.3. later in this paper.

4.4. Improvement of robustness The challenges arising from robustness issues have already been outlined in the previous chapter. Methods and concepts have to be identified that enhance existing technologies that deal with the maintenance, long-term stability and fuel flexibility of large engines. To improve robustness, components and combustion concepts optimized for wear and deposits must be made available, methods for component optimization must be developed and the engine lubrication system must be improved. The need for reliable and robust engine solutions has joined the growing demands on power output, thermal efficiency and lowest emissions as central concerns in the development of large engines. When an engine and its components are being designed, concepts such as design for reliability (DFR) are commonly used during this phase. The authors of [54] discuss this issue in detail and describe the main features of the DFR methodology. They conclude that the DFR methodology provides an effective way to create failure modebased reliability models through interrogation of Design Failure Mode & Effects Analyses and field data, which is also a big issue in [55]. Condition based engine control is mandatory to exploit the full potential of the internal combustion engine regarding power output and efficiency without violating emission and safety regulations. Possible solutions are known and have already been applied [56]. Various sensors and signals are being used for a number of control strategies [57]. All methods known so far lack the holistic optimization of the total system [58], which consists of the sensor, sensor installation and control concept. Another important factor for in-cylinder condition based engine control is the influence of operating conditions on material properties especially on sensors used for engine control. The accurate modeling of the fatigue behavior of sensors and their components is an important topic for reliable and robust engine operation. An entire model including all influencing factors on fatigue behavior of structures (e.g., roughness, residual stresses, heat treatment) is still missing, because of the huge number of relevant influences and their interactions. The contribution [59] proposes a systematic categorization of basic influencing factors on fatigue strength and life time prediction of structures including load-time history, component geometry, material properties and manufacturing process. A proper life time prediction of components and the application of optimization methods require profound knowledge about service loadings and requirements of the components and proper material properties influenced by the manufacturing process. Special effort has to be taken to enable the transferability from the laboratory testing results to the components performance [60]. In addition the manufacturing process plays an important role assessing the fatigue properties due to surface roughness and residual stresses [61]. In the case of flat specimens taken out of components or sheets the experimental procedure has to be carried out with special effort to avoid additional influences due to testing setup [62]. Besides in-cylinder condition based engine control, wear and deposit optimized combustion concepts and engine components are key. All major components of a combustion engine are exposed to wear over their life cycle. Some parts of the engine also encounter deposit formation. Both effects lead to the deterioration of the engine and component functionality. Particularly, limited spark plug life leads to additional life cycle cost. This cost can only be reduced by making use of improved ignition concepts or wear resistant spark plugs, which is one of the findings in [63]. In order to improve spark plug life the main parameters of wear have to be known. Part of the electrical energy input from the ignition system is absorbed by the electrodes. This leads to local

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

heating of the electrode surface and electrode wear, which is studied in the contribution [64]. Several different models of the electrode wear mechanism due to plasma discharge have been proposed in the past, i.e. the particle ejection model [65], the theory of sputtering [66], and erosion due to vaporization [67]. More recent publications [68,69] show that the erosion process in metals is strongly related to the volume of molten pools caused by the breakdown and the arc mode of high pressure gas discharge during the initial capacitive discharge phase of the ignition system [70]. Besides the physical properties of the electrode material, e.g., melting enthalpy and melting point, the chemical resistance to the highly oxidative atmosphere of the discharge plasma (as encountered in spark ignition discharges) is another important factor in electrode wear [71]. To operate internal combustion engines safely and efficiently, also the lubrication of the engine is of great importance. The goal of achieving reliable lubrication systems for large engines therefore is still in the main focus in the field of robust engine solutions [72]. For improved engine lubrication the optimization of the whole lubrication system needs to be addressed. Aspects like engine tribology and tribo-chemical changes of the engine oil during lifetime are still urgent problems. Methods like on-site engine oil condition monitoring and the development of new oil compositions for the use in large engines are possible solutions [73,74]. In order to make progress in the research for new lubrication systems, it is also necessary to have as precise knowledge as possible about lube oil consumption. This is particularly important from the perspective of component development (optimization of the ring/piston/liner system). Furthermore, lubricant consumption has a significant influence on the quality of the exhaust gas regarding particulate emission and the functioning of the exhaust gas aftertreatment system. In addition, increased oil consumption promotes deposit formation on the engine parts and can lead to abnormal combustion phenomena [75].

4.5. Fuel flexibility Gas engines face the additional challenge of the fluctuating quality of grid gas. In the future, the quality of grid gas will fluctuate even more because of the diversification of its sources and – more importantly – the more extensive feeding of biogas and hydrogen as well as synthetic methane from power to gas facilities. As already described in Section 3.3, the regulations of the gas quality comprise many different standards in the EU as well as in the US, which impose additional challenges to the engine manufacturers. Unfortunately most regulations lack of making reference to the calorific value, the Wobbe Index or the methane number. Thus, the change in methane number, which has a strong impact on the knock behavior of gas engines, is not yet regulated. Table 1 provides an overview of the wide variety of compositions of gaseous fuels used in combustion engines [76]. A more detailed overview of various gas compositions can be found in the results of a study [77] that highlights the aspects of gas quality for LNG (liquefied natural gas) and LBG (liquefied biogas) to be used as fuel in trucks. The composition and thus the quality of LNG are by nature highly dependent on the source. However, the methane number is influenced not only by its source but also by the liquefaction process, which can lead to different shares in long chain hydrocarbons [40]. The composition of the fuel gas determines the limits of power density and controllability in spark ignited gas engines. The limit to knocking combustion is of great importance. While higher hydrocarbons and hydrogen increase the knock tendency, inert gases lower it. The exact knowledge of the knock limits of a gas mixture is critical.

9

In the literature several criteria to determine the knock resistance of different fuel gas compositions can be found. Most of the described criteria aim on determining a methane number [78]. The methane number is determined similar to the octane number by using a specifically designed single cylinder engine. On this engine, a defined knocking combustion is first accomplished with the fuel to be evaluated. With the same engine settings, a reference fuel showing similar knocking behavior is determined; details on this procedure can be found in [79]. The AVL method for determining the methane number describes the relation of a knock resistant gas (methane) and a gas prone to knocking (hydrogen). Following the definition, a gas consisting of pure methane represents a methane number of 100. A mixture of 80 vol% of methane and 20 vol% of hydrogen represents a methane number of 80, cf. [80]. Extending the methane number to values above 100 is accomplished by admixing carbon dioxide to the methane. By evaluating different gas compositions including higher-order hydrocarbons and inert gases (nitrogen and carbon dioxide), a series of diagrams were derived to estimate the methane number. The correlation of the Southwest Research Institute is primarily used to estimate the motor octane number (MON) of a specific fuel gas mixture. Based on regression from existing measurement data, a functional depending on six gas compositions was derived. This functional considers CH4, C2H6, C3H8, C4H10 as well as CO2 and N2 [81]. The presented correlation for MON can be converted to a methane number by an additional linear correlation. However, the application of this method is limited, cf. [82]. By using interpolated AVL ternary diagrams, the Danish Gas Technology Centre derived a method for calculating the methane number. Due to the interpolation, the method lacks accuracy compared to the AVL method [83]. A recent criterion presented by the Large Engines Competence Center (LEC) [84] was developed to overcome a further shortcoming of the methane number. It has been observed that the maximum achievable IMEP (indicated mean effective pressure) of an engine operating at the knock limit can be very different depending on the gas composition, even if the compared gas compositions have the same methane numbers [80]. Therefore, a knock index has been introduced that overcomes the drawbacks of methane number by taking into account real combustion behavior. With this newly introduced index the maximum achievable engine load can be determined depending on the gas composition. One significant advantage of this index is its suitability for different engine configurations, making this method universally applicable; no specially equipped research engine is required. If the gas composition changes during engine operation, several engine parameters have to be adjusted based on the knock criterion to avoid damage to engine components from knocking combustion as well as to comply with emission limits. Furthermore, optimum efficiency and maximum possible power output must be ensured at all times. To sum up, gas engines will be able to react quickly to fluctuating gas quality if the required limits for gas characteristics are determined, knocking is avoided, and highly sophisticated engine control strategies are applied. 4.6. Engine control strategies for enhanced fuel flexibility When utilizing gases with high H2 or CO content not only the optimum efficiency but also a high uptime of the engine is crucial. As described above, additional challenges arise with the use of special gases, making comprehensive combustion control and monitoring imperative. Cylinder pressure transducers measure the pressure trace of each cylinder and calculate characteristic values, e.g., the indicated mean effective pressure (IMEP). By analyzing the measured cylinder pressure trace of each cylinder, it is possible to

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

10

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

Table 1 Typical compositions [%vol] of selected gaseous fuels for large engines [76]. Gas type

CH4

C2H4

C2H6

C3H8

C4H10

H2

CO

CO2

N2

Other

Natural gas Flare gas Sewage gas Biogas Landfill gas Wood gas Coke gas Blast furnace gas Coal seam methane

75–98 60–90 60–66 45–70 45–50 3–7 25–31 – 25–60

– – – – – 0–2 – – –

0.6–7.2 2–20 – – – – 0–1,6 – –

0.2–1.3 3–15 – – – – – – –

0.1–0.6 2–10 – – – – – – –

– – 0–3 0–1 – 6–19 54–57 2–4 –

– – – – – 9–21 5.5–8 20–30 0.1–0.4

0.1–1.6 – 32–33 25–55 35–40 11–19 1.2–2.3 20–25 1–6

0.8–9.8 – 1–5 0.01–5 9–15 42–60 3.8–9.7 45–60 4–40

1–11 – – 0–10 0–1 – 0–1 – 7–17

detect when glow ignitions occur and to identify appropriate countermeasures to avoid engine damage. A possible countermeasure to prevent damage from glow ignition is fast power reduction, which rejects the load quickly and reduces fuel gas mass immediately [39]. If fast power reduction does not work and glow ignitions still occur, the engine has to be shut down. 4.6.1. Combustion control based on gas quality When hydrogen-rich gases are used, a maximum H2 share in the fuel gas is permitted depending on the engine configuration and desired power output. Exceeding this limit will result in an engine stop. Along with the maximum value, the gradient of the fluctuation of H2 in the mixture has a great impact on stable engine operation and thus is also considered in the control strategy. While conventional engine control can compensate for a decrease in H2 content, an increase in H2 content represents a danger to the engine due to subsequent accelerated combustion and possible glow ignition. If the H2 content suddenly increases beyond a critical value, the engine performs a quick load reduction. After the combustion is adjusted to the higher H2 level by primarily setting an adequate ignition timing, the engine recovers the original power output. The scenario of rapidly increasing H2 content often occurs with steel gas applications [39]. 4.6.2. Model based combustion control During engine operation, various boundary conditions such as gas composition, mixture supply, inlet and outlet valve condition and deposits in the combustion chamber change over time and thus influence the combustion in gas engines. Model based combustion control can be applied in order to attain a high efficiency and high uptime of the engine especially when non-natural gas is used. These models are based on either combustion phasing or combustion stability. Combustion control is particularly crucial when hydrogen rich gases are used. Slow energy conversion can lead to high hydrogen content in the exhaust, thereby compromising emissions or even leading to combustion in the exhaust. In contrast, rapid energy conversion can lead to engine damage. By applying cylinder selective ignition timing control, the combustion phasing can be kept within the desired range for each cylinder. In the case of knocking, combustion control can adjust the ignition timing accordingly. Applying this control strategy leads to robust engine operation with which also fluctuations in gas quality can be compensated. Combustion stability can also be improved based on the coefficient of variation of the combustion. If this value deviates from a defined reference value, the gas share in the mixture is adjusted according to the operating point. To sum up, combustion control based on the coefficient of variation has an impact on all cylinders while combustion phasing control has a separate effect on each individual cylinder. A combination of both control concepts enables high power density and high robustness of the engine.

5. Development methodology The previous sections discussed the challenges in technology development presented by the requirements of sustainable power generation. The following section presents a methodology able to meet all current and future requirements for developing combustion systems and engine components. 5.1. State-of-the-art methodology The state-of-the-art LEC Development Methodology (LDM) was elaborated in order to support efficient engine development. Papers [6,85,86] are examples of the successful application of this method in the development of large gas engines at the LEC. In recent years, many research institutions and engine manufacturers have employed this method, cf. [87–89], or apply a similar method in their engine development process. As a result, this method can be regarded as the standard in the area of large engines. LDM is based on the intensive interaction between simulation and experimental investigations on single cylinder research engines (SCE), see Fig. 5. To develop and optimize combustion concepts, the methodology makes use of 0D and 1D engine cycle simulation as well as 3D CFD simulation. 0D/1D engine cycle simulation is applied to preoptimize significant engine parameters such as compression ratio or valve timing. Moreover, the burn rate curve can be precalculated in this phase of development by applying quick 0D models [90]. Specially designed models are able to deal with non-natural gases in general, cf. [91], or syngas, cf. [92]. These models can even be extended to define an index that sets the load that can be achieved with a certain fuel gas in relation to a baseline point with natural gas [84]. 3D CFD simulation is employed above all to optimize the details of related processes, e.g., mixture formation and combustion in the prechamber and main combustion chamber and determination of the location of knock. Both simulation and experiments may also rely on statistical methods such as the DoE method, as has been shown in [6,86]. Subsequently, the results from single cylinder engine tests must be transferred to the multicylinder engine (MCE). To this end, it must be ensured that the thermal boundary conditions as well as the conditions at the beginning of the intake stroke (temperature, pressure, and working gas composition) on the single cylinder engine are comparable to those of the multicylinder engine. An iterative process based on 1D engine cycle simulation provides these conditions. 5.2. Advanced methodology 5.2.1. General approach To meet the challenges and technological development requirements of power generation with renewable fuels, new paths must be taken to improve applied technologies and the development

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

11

Fig. 5. State-of-the-art development methodology (LDM).

Fig. 6. Transition from LDM to LDM Advanced.

methodology for gas engines. While LDM currently focuses on the development of steady-state combustion concepts, a transition to more integrated treatment of all combustion related processes that requires a significantly more comprehensive approach will be key in the future. LDM Advanced, currently in use at the LEC, considers also combustion related aspects such as durability, wear, ignition, and fuel supply as well as the development of transient combustion concepts and controls, see Fig. 6. LDM Advanced requires the detailed physical modeling of all engine components. For example, detailed models to describe spark initiation and removal of material along with parameters derived from thermodynamic simulation must be made available and linked to an overall spark plug wear model. The integration of expertise from a variety of disciplines is required to accomplish this task. Furthermore, control systems incorporate an example of this comprehensive approach. The complete process from the control concept itself to the optimal integration of the sensor into the cylinder head as well as durability issues and algorithms for recognizing sensor errors have to be considered. Because the demands placed on engine development continue to increase, automated methods for analysis and plausibility

checks of measurement data are necessary in order to ensure the highest possible quality of test results [93]. Since very high thermal efficiency for large engines has already been obtained, further engine development all too often results in only small improvements in efficiency and performance values. To assess even small steps in the technology progress correctly, it is mandatory to develop advanced methods that lead to correct conclusions even though the measurement technique has reached the limits of accuracy. The main research challenge is to develop a fully integrated methodology for performing online plausibility checks of measurement data and to derive confidence intervals for measurement results and elaborate routines for automatic failure correction directly on the test bed. The automatic procedures for improvement of measurement results are of utmost importance for LDM to work and have been researched and published in recent years. The basics of the procedures are presented in [94]. This work investigates processes for validating and checking the plausibility of measured data from the test bed. Gaussian correction calculations are applied and it is claimed that the corrected values must fall within the boundaries of the measurement accuracy of the corresponding measuring

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

12

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

Fig. 7. Main aspects addressed by LDM Advanced.

device as well as fulfill the physical conditions between the input values. The authors conclude that with this methodology also the accuracy of derived values (i.e. values that cannot be measured directly at the test bed) can be assessed. Based on these first steps further development was conducted [95]. In this contribution a variant of the methodology is set up by introducing an improved calculation of thresholds that enable highly sensitive fault diagnosis and robustness at the same time. The performance of the methodology is evaluated using simulated faults. Nearly 80% of faults were detected when fault intensities were higher than 10% and almost every detected fault was able to be isolated correctly. In [96] the methodology is implemented using a modular structure, i.e. the applicability to various scenarios is enabled. The modules comprise a low number of system parameters and high universality which enables the fault diagnosis method to be adapted to a new test engine with only small configuration effort. 5.2.2. Main innovations of LDM During the transition from LDM to LDM Advanced, the development goals for gas engines were switched from topics related to performance to topics related mainly to robustness, cf. Figs. 4 and 6. The challenges and requirements for these development goals have already been discussed earlier in this paper. Fig. 7 provides an overview of the main aspects of LDM Advanced and shows what engine features in the areas of performance and robustness are addressed in order to achieve the said goals, which can be regarded as the primary development goals for the gas engine applications of the future, i.e. load response, maintenance, long term stability and fuel flexibility. A transient SCE test bed is applied to develop strategies for achieving the required load response goals. Wear assessment strategies cover the areas of durability, oil consumption, oil quality as well as the effect of wear mechanisms on various engine components, e.g., piston, prechamber, spark plug. Depending on the area of application these strategies can comprise experimental investigations, the use of simulation models as well as a combination of both methods. As an example, the assessment of spark plug wear can be stated. Using these strategies, maintenance and long-term stability can be improved. Finally, strategies for reactions to various gas qualities have to be provided that can adapt the engine parameters to fluctuations in gas quality and enable fuel flexibility in large gas engines. These strategies can be based on measured values from cylinder pressures or knock sensors and can be used to adjust the engine parameters. In general, LDM Advanced mainly focuses on development processes that are performed without the need for extensive testing on a multicylinder engine. These processes rely primarily on simulation tools as well as measurements on a SCE. However, experiments on special test rigs for fundamental investigations are also included in order to support simulation model development. In the field of combustion concepts for NNG in particular,

development cannot rely on extensive testing on a multicylinder engine and must be done in a way that the concept can be directly transferred to the on-site facility. The following sections presents three examples of how the methodology has been applied in gas engine development. The demands of transient engine operation development can be met by elaborating a Hardware-in-the-Loop (HiL) setup for large gas engines on single cylinder test beds; more details about this setup will be presented in the next section. This setup enables the development of control concepts for large engines on an SCE in the absence of a multicylinder engine. To predict spark plug wear, the wear mechanisms as well as the parameters that influence spark plug wear must be thoroughly understood and adequate wear models have to be identified. The input for these wear models is generated by thermodynamic simulation models able to simulate ignition behavior and flame initiation in a large engine under various boundary conditions, which is explained in detail below. In the field of large engine applications fueled by NNG, tailormade engine solutions must be developed that allow efficient, low emission use of each gas. At the same time, the cost of development of combustion concept design and optimization must be kept down in order to remain economically viable. Strategies for reactions to various gas qualities have to be elaborated in order to guarantee enhanced fuel flexibility and robust engine operation even for applications with high fluctuations in gas quality. I.e. the engine has to react to different gas compositions by adjusting its operating parameters. As an example for a tailor-made engine solution, the development of a combustion concept for blast furnace gas applications is shown. 5.2.3. HiL operation of an SCE to optimize the transient behavior of large gas engines As previously mentioned in this paper, it is critical to improve the transient characteristics of gas and dual fuel engines so they are able to compete with diesel engines in the area of power generation. Testing of transient behavior as well as design of different control strategies for the multicylinder engine (MCE) should be carried out on a SCE test bed during the development process in order to avoid the expenses associated with development on a MCE. LDM Advanced accomplishes this by using a method that includes both simulation and measurements. Fig. 8 presents the basic layout of a HiL setup [53] of a large two-stage turbocharged gas engine. The main components of the HiL setup are a SCE test bed that includes a SCE, transient SCE test bed controls and a real-time controller. Simulation models used in HiL mode are designed and calibrated using measurement data from the MCE or from calibrated MCE simulation models in 1D simulation programs. These models must provide the required values, i.e. pressures and engine speed, for transient control of the single cylinder test bed in real-time (RT models). Besides controlling

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

13

Fig. 8. Layout of the HiL setup for transient SCE operation.

the SCE test bed, the real-time controller serves as a platform for the models and a simplified MCE control unit. Fig. 9 provides a detailed overview of the HiL layout as it is applied to transient SCE operation in island mode. In contrast to grid-parallel mode where the engine speed is determined by the power grid, island mode also requires models that describe the transient behavior of the powertrain (i.e. speed changes) during load changes. The three main components of this design are the simplified MCE engine control unit (MCE-CU, cf. red box in Fig. 9), the realtime models (blue1 boxes) and the SCE test bed (green box). The simplified MCE-CU consists of three main control algorithms:  Knock control  Engine speed control and skip firing strategy  NOx or excess air ratio control. Knock control shifts the ignition towards late crank angles when knocking combustion is detected. This is an important feature for gas engines because of the very narrow operating range between misfire and knocking combustion. Engine speed control is used for island operation when the speed is not determined by the power grid. When the power demanded by consumers does not match the power provided by the engine, the engine speed changes accordingly. In the event of a deviation in engine speed, the MCE-CU adjusts the amount of gas by changing the port injection opening duration. At low loads, a skip firing strategy comparable to that of the MCE must be implemented. The occurrence of skip firing is an important input for the real-time model that describes the gas exchange since the turbochargers are influenced by the exhaust gas enthalpy. The information about skip firing is also sent to the powertrain model so that the indicated power for skipped cylinders is adequately reduced. The NOx or excess air ratio (EAR) control leans or enriches the mixture according to the given target value. The controller output adjusts the amount of charge air blow off, which directly influences the charge air pressure of the virtual MCE. This calculated charge air pressure is the target value that is set on the SCE test bed by the SCE-CU. The combustion EAR and subsequently the 1 For interpretation of color in Fig. 9, the reader is referred to the web version of this article.

NOx emissions are adjusted in response to the charge air pressure of the SCE. Three main real-time model modules have been defined (cf. the content of the two light blue boxes in Fig. 9), namely the physicsbased gas exchange model, the physics-based powertrain model and the functional relationships for main combustion chamber (MCC) gas supply pressure and pre-chamber (PC) gas supply pressure. The last of the three are referred to as non-physical models; these gas pressures are simply functions of the current charge air pressure. The measured SCE cylinder pressure at exhaust valve opening (EVO), the engine speed, a skip firing flag and a value describing the compressor bypass mass provide the input to the gas exchange model. The transient charge air pressure trace and the exhaust back pressure trace are the output variables from the gas exchange model to the SCE-CU. The skip firing flag, the indicated SCE power and a target consumer power serve as the input to the powertrain model. The engine speed set on the test bed is one output of the powertrain model to the SCE-CU. In this approach, the gas exchange model is the most important real-time model. The gas exchange of the MCE including two-stage turbocharging is described by a 0D model that is implemented in Matlab/Simulink by setting up various elements for volumes, turbochargers, cylinders, etc. These elements contain the required calculation formulas and are interconnected according to the engine layout. The pipes and the cylinders are described in 0D using the comparably simple and fast fill-and-drain method. To achieve real-time capability, the 1D effects in the piping system as well as detailed modeling of the gas supply system including possible pulsation effects are not examined. The validity of the real-time models was assessed by comparing them with the results of 1D flow modeling in GT-Power as well as MCE measurement data. Although there were deviations between simplified real-time modeling and comprehensive 1D calculation, the quality of the agreement was sufficient for control purposes. The model was implemented numerically into the real-time controller; it works stably and quickly enough to ensure real-time capability for closed loop operation on the test bed. The upgraded test bed can now be used as the staging ground for developing transient MCE control strategies, testing components under transient operation conditions and continuing combustion development for transient operation.

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

14

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

Fig. 9. Detailed overview of the HiL layout (for island mode).

A thorough study in [53] provides detailed information on the application and the verification of the models as well as comparisons of real-time model results to measurement data from transient MCE operation including load steps and load ramps. 5.2.4. Electric arc modelling in spark ignition models for large gas engines In order to increase long-term stability of large gas engines, spark plug wear is an important issue that must be addressed. Current design approaches to developing spark plugs are incapable of properly describing all the influencing factors, especially under real engine conditions. The main challenge lies in correctly describing the conditions that lead to wear. In advanced design of spark plugs, it is imperative to completely understand these factors and their interactions. Since the conditions generated by the ignition process significantly affect spark plug wear, thermodynamic models of the ignition process are required. A detailed understanding of the physics of arc breakdown and flame kernel initiation is also necessary in order to ensure stable ignition under increasingly severe engine conditions. With this knowledge, adequate models are formulated that can then be implemented into simulation tools for application, e.g., in the optimization of ignition performance. In

line with LDM, the modeling process is accompanied by fundamental experiments to ensure its correctness. Advanced ignition models require a detailed description of the mutual interaction between the electronic ignition system, the electric arc and flame initiation. Thorough simulation of breakdown phenomena where the first conducting arc channel is formed between the spark plug electrodes is hardly feasible as part of internal combustion engine CFD simulation because it involves scales of time and length that span several orders of magnitude. Consequently, current CFD simulation of internal combustion engines relies on comparably simple sub-models of different physical complexity to initiate the combustion model. Several recent papers present various comprehensive spark ignition models for CFD simulation that take into account the behavior of an electric arc under local flow conditions at the spark plug [97–100]. Because the models from the literature were designed primarily for gasoline engines, these ignition models cannot be transferred directly into CFD simulation of large stationary gas engines. In these engines, ignition is initiated at significantly higher pressures and the spark plugs are characterized by electrode gaps that are much smaller than those in automotive engines. In contrast to the models described in the literature in which the electric arc passively follows the local flow, the

Fig. 10. Improved arc model in detailed spark ignition models.

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

Fig. 11. Comparison of simulated and measured arc length over time.

improved arc model also accounts for the relative movement between the electric arc (region with a high temperature) and cold flow. Fig. 10 provides an overview of this methodology as well as the concept for an arc model. More comprehensive information on this model can be found in [101,102]. In the CFD code, the user defined function (UDF) that contains the electric arc model considers the influence of local flow velocities on the energy transferred from the ignition system to any potential infant flame kernels. These kernels define the initial condition for the combustion model of the main CFD solver. In order to compute the local flow in the vicinity of the electrode surfaces, the spark plug has to be included in the geometry of the combustion chamber. In CFD simulation, the arc is represented by marker particles placed along a line between the electrodes. The electric arc is seen as electrical impedance in the electric network that represents the (capacitive) ignition system. The length of the arc determines the electric resistance and thus the discharge characteristic of the ignition system. Arc resistance was modeled with an empirical relationship found in the literature [97].

15

In contrast to prior model approaches in which passive movement of the arc with the local flow yields the elongation of the arc over time, this method relies on an advanced approach that makes use of an arc column model to consider two contributions to arc movement. First, the arc column is regarded as an electric conductor. As a result, the electric current through the curved conductor exerts a self-induced magnetic field on itself; this field is described by the electromagnetic forces in the momentum equation. Second, the current flowing through a curved conductor leads to inhomogeneous heating, resulting in a relative movement of the arc plasma in the direction of the curvature because of the contribution of Joule heating in the energy equation. An order of magnitude analysis of these effects suggested that inhomogeneous arc heating is the dominating effect for typical arc currents; a model of the relative velocity was proposed in [101]. Fig. 11 provides an example of a comparison of the computed temporal evolution of arc lengths with evaluated values from high-speed images with different flow velocities obtained in a test rig at 1 bar [101]. The dashed line in Fig. 11 indicates the result obtained when the sub-model for relative velocity is ignored, which leads to a significant overprediction of the arc movement. To verify the arc model at conditions more relevant to ignition in gas engines, the LEC is currently performing experiments in a high pressure flow test rig with an optical access. Pressures from 1 to 60 bar and flow velocities of 0 to 30 m/s can be obtained and will provide a good basis for model validation. The arc model presented in this section is currently being integrated into a comprehensive ignition model for large gas engines that is under development at the LEC. 5.2.5. Combustion concept for blast furnace gas applications The following section presents a sophisticated concept for sustainable power generation in large gas engines that efficiently and flexibly exploits blast furnace gas (BFG). The objective was to develop a highly efficient, high power output combustion concept for a large gas engine in the 4 MW power range using BFG [103]. This concept must adhere to the TA Luft emission limits [10], ensure sufficient combustion stability despite fluctuations in gas quality and exhibit good starting behavior. The development approach was derived using the advanced LDM.

Fig. 12. Predesign of the combustion concept using 3D CFD simulation.

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

16

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

Fig. 13. Operating range in relation to H2 content.

In the first step, a suitable concept for the combustion of BFG was selected and preoptimized using simulation tools. The extremely unfavorable properties of BFG, which contains a great amount of inert gases and possesses a calorific value lower than 1 kWh/m3n, require a very specific combustion concept. Gas scavenged, mixture scavenged and unscavenged prechamber concepts as well as open chamber concepts are all possible with large gas engines. 0D/1D simulation revealed that with the prechamber concepts that are normally used with this engine size, ignition cannot be successfully induced due to the unfavorable mixture composition in the prechamber at ignition timing. This information was obtained by determining the mixture composition in the prechamber at ignition timing with a 1D simulation model that also provides a detailed description of the mass overflow between the prechamber and main combustion chamber. Thus, the main focus of development was on the open chamber concept. With this concept, a sufficient level of turbulence induced by charge motion must be achieved to obtain an appropriate flame speed. To obtain this level of turbulence in the combustion chamber during combustion, a very high swirl level was combined with an appropriate piston shape. 3D CFD simulation was extensively used during the optimization process. Fig. 12 presents selected results from this process, whereby the locally averaged turbulent kinetic energy (TKE) was chosen as a criterion to evaluate the different variants. The chart shows the optimized variant as well as two intermediary development steps (variants A and B) in relation to the baseline variant. Based on these investigations, the most promising piston shapes in combination with the swirl level were selected for the experimental tests. Measurements were taken on a single cylinder research engine to verify the simulation results and to determine certain engine performance values (e.g., efficiency, emissions level, stability). Combustion chamber variants with increased TKE shortened combustion duration, which resulted in a significantly faster fuel conversion with the optimized variant. After the best piston variant had been chosen and the optimal compression ratio determined, the engine operating parameters (e.g., ignition timing, mixture temperature, charge pressure) were optimized. The goal was to obtain the largest operating range possible. The fluctuations in the composition of BFG cause changes in the combustion behavior of the gas. The share of hydrogen (H2) in particular noticeably influences combustion. Thus investigations were carried out on the single cylinder research engine to evaluate the sensitivity of the engine concept to fluctuations in H2 content. Fig. 13 presents the results from these investigations with the

optimal variant; the misfire limit and the knock limit restrict the operation range. A special control concept adapted to BFG application was developed for the multicylinder engine in order to ensure stable engine operation with variable load and fluctuating gas quality. Through continuous optimization of both combustion and gas mixing, a maximum energy yield can be guaranteed at any time and at any operating point. The simulation based control concept relies upon on-board cylinder pressure measurement and admixing of an additional gas (e.g., coke gas). The concept was optimized on a prototype engine in a French steel mill. 6. Summary and conclusions This paper has provided an overview of the technological challenges as well as the demands placed on the next generation of gas engines along the road to sustainable power generation. Highly sophisticated technology must be developed to meet these challenges and to deal with the growing share of renewable power sources in the overall energy supply. The most important challenges are still the basic need to use available fossil fuels as efficiently as possible while complying with increasingly stricter emission limits. Synthetic fuels from renewable sources as well as waste gases are growing in popularity and present additional challenges in all engine applications. This paper identifies technologies required to meet these challenges, namely improvements in robustness and dynamic behavior that will allow gas engines to compete with diesel engines in applications with high transient requirements. Combustion systems and control systems that meet the high demands on both transient behavior and knock resistance of the engine are required to deal with the great fluctuations in gas quality anticipated with grid gas and LNG. Technologies that enhance fuel flexibility without compromising efficiency and emissions will continue to increase in importance. A development methodology was presented that enables the development of combustion concepts to meet these technological requirements. This methodology allows direct implementation of an engine concept into the on-site multicylinder engine without prior cost-intensive tests on the multicylinder engine. The reduced development cost and time for the engine manufacturer can be passed on to the user and is thus an important enabler of this technology that aims to produce energy with a low environmental impact and few CO2 emissions. The main aspects of the methodology – the simulation of transient MCE behavior on an SCE test bed and the development of

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

advanced ignition concepts and combustion concepts for enhanced fuel flexibility and reduce wear – can be regarded as the primary development goals for the gas engine applications of the future. A significant part of optimization work can be carried out in advance by making extensive use of simulation to evaluate different engine configurations.

Acknowledgements The authors would like to acknowledge the financial support of the ‘‘COMET - Competence Centres for Excellent Technologies Programme” of the Austrian Federal Ministry for Transport, Innovation and Technology (BMVIT), the Austrian Federal Ministry of Science, Research and Economy (BMWFW) and the Provinces of Styria, Tyrol and Vienna for the K1-Centre LEC EvoLET. The COMET Programme is managed by the Austrian Research Promotion Agency (FFG).

References [1] Budzianowski et al. Perspectives for Low-carbon Electricity Production Until 2030: Lessons learned from the comparison of local contexts in Poland and Portugal. Energy Sources - B 2016;11(6):534–41. http://dx.doi.org/10.1080/ 15567249.2012.710293. [2] Warnecke W, Karanikas J, Levell B, Mesters C, Schreckenberg J, Adolf J. Gas – Eine Brückentechnologie für zukünftige Mobilität? In: 34. Internationales Wiener Motorensymposium, 25–26 April 2013. [3] Zauner S, Arnold G, Kopecek H, Kumar C, Spreitzer C, Trapp C, et al. Nutzung von Gichtgas im Großgasmotor mit Hilfe zylinderdruckbasierter Motorregelung und Zweigasbetrieb. In: 3. Rostocker Großmotorentagung ‘‘Die Zukunft der Großmotoren III – Großmotoren im Spannungsfeld von Emissionen, Kraftstoffen und Kosten, 18–19 September 2014. [4] Wimmer A, Schneßl E, Simon N. HyWood - Nachhaltige Strom- und Wärmeerzeugung mit wasserstoffreichem Synthesegas aus Biomasse. In: Submission and nomination for the Energy Globe Styria Award, Graz; 2016 [5] European Council Conclusions. 2030 climate & energy framework, Documentation of Conclusions - 23/24 October 2014. [6] Wimmer A, Winter H, Schnessl E, Pirker G, Dimitrov D. Brennverfahrensentwicklung für die nächste Gasmotorengeneration von GE Jenbacher. In: Presented at the 7th Dessau gas engine conference, Dessau, Germany, March 24–25, 2011. [7] Grotz M, Böwing R, Lang J, Thalhauser J, Christiner P, Wimmer A. Efficiency Increase of a high performance gas engine for distributed power generation. In: Presented at CIMAC Cascades, Graz, 27/02/2015. [8] Gothenburg Protocol 2010. Protocol to the 1979 convention on long-range transboundary air pollution to abate acidification, eutrophication and ground-level ozone, Gothenburg, 1999, available in the internet, , Status: 30.07.2010. [9] Council of the European Union. Interinstitutional File: 2007/0286 (COD), Brüssel, 2009, Annex V S. 14. [10] Federal Ministry for Environment. Nature conservation and nuclear safety. Technical instructions on air quality control – TA Luft of 24 July 2002. [11] Tinschmann G, Birgel A, Trapp C, Schnessl E, Redtenbacher C, Wimmer A. Large gas engines – 75 mg/Nm3 @ 15%O2 nitric oxide ‘Engine-internal measures or exhaust aftertreatment?’. In: CIMAC proceedings, Paper No. 296, Shanghai 2013. [12] Schinagl G. Innermotorische Maßnahmen zur Reduktion der Formaldehydemissionen von Großgasmotoren, Dissertation, Graz University of Technology, 2010. [13] CIMAC Working Group ‘‘Gas Engines”. Position Paper ‘‘Methane and Formaldehyde Emissions of Gas Engines”. Conseil International des Machines A Combustion, April 2014. [14] EUROMOT 2012. Euromot Position ‘‘Methane Slip from Internal Combustion Gas Engines”, The European Association of Internal Combustion Engine Manufacturers - EUROMOT, 27. April 2012. [15] Wimmer A, Pirker G, Engelmayer M, Schnessl E. Gas engine versus diesel engine: a comparison of efficiency. MTZ Ind 2011;11:2–6. [16] Behr T, Kahi M, Reichl A, Hubacher M. Second generation of two-stage turbocharging Power2 systems for medium speed gas and diesel engines. In: CIMAC proceedings, Paper No. 134, Shanghai, 2013. [17] Mathey C. Valve control management - the possibility of improving gas engine performance. In: CIMAC proceedings, Paper No. 389, Shanghai, 2013. [18] Trapp C, Böwing R, Birgel A, Kopecek H, Madl W, Fahringer A, et al. ‘‘GE’s J920 gas engine – 10.3 MW power at up to 50% electrical efficiency”. In: Conference proceedings, 10th Dessau gas engine conference, Dessau, 2017. [19] CIMAC Working Group ‘‘Gas Engines”. Position Paper ‘‘Transient Response Behaviour of Gas Engines”, Conseil International des Machines A Combustion, April 2011.

17

[20] ISO 8528–5 (Technical Regulation). Reciprocating internal combustion engine driven alternating current generating sets – Part 5: Generating sets. Switzerland: International Organization for Standardization, 2013–3. [21] DIN VDE 0100–710 (VDE 0100–710) (Technical Regulation). ‘‘Errichten von Niederspannungsanlagen – Teil 7–710: Anforderungen für Betriebsstätten, Räume und Anlagen besonderer Art – Medizinisch genutzte Bereiche (IEC 60364-7-710:2002, modifiziert), Deutsche Übernahme HD 60364-7710:2012”, DIN Deutsches Institut für Normung e.V. und VDE Verband der Elektrotechnik Elektronik Informationstechnik e.V., Berlin, 2012–10. [22] National Fire Protection Association. NFPA 110: standard for emergency and standby power systems 2016 edition, 2016. [23] Nuormala K. Wärtsilä 50SG gas engine for smart power generation. In: Roßlau WTZ, editor. Proceedings: 8th Dessau gas engine conference, Dessau. p. 23–41. [24] Henn R, Stellwagen K, Drexel Ch, et al. Optimisation of supercharged lean mixture gas otto engines for independent operation. In: CIMAC proceedings, Paper No. 265, Vienna 2007. [25] European network of transmission system operators for electricity (ENTSOE). ENTSO-E network code for requirements for grid connection applicable to all generators, Brussels, 2013–3. [26] Thalhauser J, Schaumberger H, Schäffert P, et al. ‘‘Betriebsstrategien für einen aufgeladenen Gasmotor bei dynamischem Netzspannungseinbruch‘‘, Contribution to: ‘‘18. Aufladetechnische Konferenz”, Dresden, 2013. [27] BDEW Bundesverband der Energie- und Wasserwirtschaft e.V. Technische Richtline Erzeugungsanlagen am Mittelspannungsnetz: Richtlinie für Anschluss und Parallelbetrieb von Erzeugungsanlagen am Mittelspannungsnetz, Berlin, 2008–6. [28] Ennemoser A, Mahmoud K, Winklhofer E. Kombinierte Fluid- und Strukturberechnung zur thermischen Analyse in Zylinderköpfen von Verbrennungskraftmaschinen. MTZ Motortechnische Zeitschrift 1999;60. [29] Dharan RB, Goud R, Manoharan R, Dhiman V. Design and development of cylinder block for high power density diesel engine using CAE/CFD tools for a tractor engine with integrated approach. SAE Technical Paper 2013–01-2753, 2013. [30] Porpatham E, Ramesh A, Nagalingam B. Effect of swirl on the performance and combustion of a biogas fuelled spark ignition engine. Energy Convers Manage 2013;76:463–71. [31] European Standard EN 16726 – Gas infrastructure – Quality of gas – Group H: [Cited 15/03/2016]. [32] EASEE-gas. common business practice ‘‘harmonisation of natural gas quality”, 2005–001/02. [33] EUROMOT 2013. ‘‘Gas Quality Harmonisation”, The European Association of Internal Combustion Engine Manufacturers - EUROMOT, 18. Septemper 2013. [34] Amplatz E, Schneider M, Trapp C. ‘‘Utilization of special gases in stationary gas engines”, 6. Internationale MTZ-Fachtagung, 15. und 16. November 2011, Kiel. [35] Ernst A, Sander U, Kibele C, Odermatt M, Wasgindt A. The new MTU series 4000 for non-natural-gas applications. In: Conference proceedings, 7th Dessau gas engine conference, Dessau, 2011. [36] Sander U, Ernst A, Kibele C, Odermatt M, Wasgindt A. Die neue MTU Baureihe 4000 für Biogasanwendungen, MTZ-Konferenz Heavy Duty- On- und OffHighway-Motoren, Mannheim, 2010. [37] Farina MF. Flare gas reduction - recent global trends and policy considerations. GE Energy | GEA18592 (10/2010), 2010. [38] Schneider M. Utilization of metallurgical process gases with gas engines. In: 28th international ferro-alloys conference Berlin, Germany, 2012. [39] Prankl S, Trapp C, Böwing R, Zuschnig A, Schiestl S, Schneßl E, Wimmer A. Fortschritte bei Sondergasen - GE Gasmotoren mit hoher Leistung für wasserstoff- und kohlenmonoxidreiche Gase. In: ‘‘Die Zukunft der Großmotoren IV”. Hrsg. Horst Harndorf. 227–243, ISBN 978-3-941554-13-9, S. [40] Schnessl E, Wimmer A. Trends in the development of large gas engines for power generation. In: REMOO 2016 – international conference & workshop – science and engineering for reliable energy, Budva, Montenegro, 2016. [41] Korb B, Wachtmeister G, Kuppa K, Dinkelacker F. THC-emissions in gas engines - experimental and numerical studies. In: 15th conference ‘‘the working process of the internal combustion engine”, Graz, 2015. [42] Ratzke A, Hennecke C, Dinkelacker F. Simulating the combustion and nearwall flame extinction of a methane gas IC engine by employing a zonal cylinder model. In: CIMAC proceedings, Paper No. 102, Shanghai, 2013. [43] Moser M, Kovacic G, Knafl A, Stiesch G. Transient operation behavior of large bore gas engines: major influences, challenges and solutions. In: 8. International MTZ conference heavy-duty, on- and off-highway engines, Ludwigsburg, DE, Nov 5–6, 2013. [44] Jubany Arribas E, Becchio D, Thalhauser J, Douglas K, Lang J, Tinschmann G. Next generation turbocharging concepts for large gas engines. In: 15th conference ‘‘the working process of the internal combustion engine”, Graz, 2015. [45] Lopez F, Graus M, Spyra N, Schaumberger H. Improved transient performance of GE’s Jenbacher type 6 gas engines. In: MTZ-Fachtagung Heavy-Duty-, Onund Off-Highway-Motoren - Speyer, Germany, 2015 [46] Kopecek H, Spyra N, Birgel A, Spreitzer K, Trapp C. Cylinder pressure based controls for robust operation of gas engines of high power density. In: 9th Dessau gas engine conference, Dessau, 2015. [47] Isermann R, Schaffnit J, Sinsel S. Hardware-in-the-loop simulation for the design and testing of engine-control systems. Control Eng Pract 1999;7:643–53. Elsevier Science.

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023

18

G. Pirker, A. Wimmer / Energy Conversion and Management xxx (2017) xxx–xxx

[48] Lahti JL, Moskwa JJ. A transient test system for single cylinder research engines with real time simulation of multi-cylinder crankshaft and intake manifold dynamics, SAE Technical Paper 2004–01-0305. [49] Snyder ME, Lahti JL, Moskwa JJ. Simulating transient multi-cylinder engine gas exchange dynamics on a single-cylinder research engine, SAE-Paper 2006-01-1074, 2006. [50] Gambarotta A, Ruggiero A, Sciolla M, Lucchetti G. HiL/SiL-System zur Simulation von aufgeladenen Dieselmotoren, Motortechnische Zeitschrift MTZ 73, 02/2012. [51] Friedrich I, Buchwald R, Stölting E, Sommer A. Das virtuelle Fahrzeug Transiente Simulation für den dieselmotorischen Entwicklungsprozess, Motortechnische Zeitschrift MTZ 70, 12/2009. [52] Wu H, Li M. A Hardware-in-the-Loop (HIL) Bench Test of a GT-Power Fast Running Model for Rapid Control Prototyping (RCP) Verification, SAE Technical Paper 2016–01-0549, 2016, doi:http://dx.doi.org/10.4271/2016-01-0549. [53] Mayr Ph, Pirker G, Krenn M, Wimmer A. Simulation-based control of transient SCE operation, SAE 2017-01-0544. [54] Calvert I, et al. Integrated design, analysis and development processes applied to the design of a high specific output gas engine cylinder head. In: CIMAC proceedings, Paper No. 419, Shanghai 2013. [55] Tinga T. Principles of loads and failure mechanisms, applications in maintenance, reliability and design. Springer series in reliability engineering 2013, Springer Link, ISSN: 1614–7839. [56] Ritscher B, Greve M. CATERPILLAR M46 Dual Fuel engine with new cylinder pressure based control strategies. In: CIMAC proceedings, Paper No. 411, Shanghai, 2013. [57] Henschen P. ,Entwicklung einer zylinderdruckbasierten Regelung für mittelschnelllaufende Gasmotoren. In: 8th Dessau gas engine conference, Dessau, 2013. [58] Rehbichler G, Kendlbacher C, Bernhaupt M. The bosch electronic diesel control system for medium and high speed engines. In: CIMAC proceedings, Paper No. 141, Shanghai 2013. [59] Eichlseder W. Transferability specimen – component: influences on the fatigue strength. In: FemFat user meeting 2011, Steyr, 2011. [60] Maderbacher H, Gänser H-P, Riedler M, Stoschka M, Stockinger M, Eichlseder W. Lifetime optimization of hot forged aerospace components by linking microstructural evolution and fatigue behaviour. Adv Mater 2011;278:162–7. [61] Javidi A. Influence of machining on the surface integrity and fatigue strength of 34CrNiMo6 steel [Doctoral Thesis], Montanuniversiät Leoben, Austria, 2008. [62] Strohhäussl B, Winter G. Influence of elevated temperatures on high cycle fatigue behaviour of thin sheet metals. In: 7th international conference on low cycle fatigue, Aachen, Germany, 2013, p. 371–6. [63] Trapp Ch, Kraus M, Laiminger S. Ignition concepts of modern gas engines – a foresight. In: 12th conference ‘‘the working process of the internal combustion engine”, Graz, 2009. [64] Jeanvoine N, Muecklich F. FEM Simulation of the temperature distribution and power density at platinum cathode craters caused by high voltage ignition discharges. J Phys D Appl Phys 2009;42:035203. [65] Gray EW, Pharney JR. Electrode erosion by particle ejection in lowcurrent arcs. J Appl Phys 1974;45:667–71. [66] Sigmund P. Theory of sputtering. Phys Rev 1969;184:383–416. [67] Jones FL. Electrode erosion by spark discharges. Br J Appl Phys 1950;1:60–5. [68] Lasagni A, Soldera F, Mücklich F. FEM simulation of local heating and melting during electrical discharge plasma impact. Modell Simul Mater Sci Eng 2004;12:835–44. [69] Soldera F, Mucklich F, Hrastnik K, Kaiser T. Description of the discharge process in spark plugs and its correlation with the electrode erosion patterns. IEEE Trans Veh Technol 2004;53:1257–65. [70] Fridman A, Kennedy LA. Plasma physics and engineering. CRC, 2011. [71] Rager J. Funkenerosion an Zündkerzenelektroden [Ph.D. thesis]. Universität des Saarlandes, 2006. [72] Rea S. Development of new gas engine oils for superior corrosion and deposit control in severe gas applications. In: CIMAC proceedings, Paper No. 180, Bergen 2010. [73] Schasfoort Th, Busse P, Holdsworth R. Development of a special lubricant for gas engines running on sour gas. In: Conference proceedings, 7th Dessau gas engine conference, Dessau, 2011. [74] Novotny-Farkas F, Kaltenböck D, Schmied C, Simmerer B. Oil condition monitoring of biogas-powered stationary engines – methods and concepts. In: 12th conference ‘‘the working process of the internal combustion engine”, Graz, 2009. [75] Yasueda S, Takasaki K. The abnormal combustion caused by lubricating oil on high BMEP gas engines. I: 13th conference ‘‘the working process of the internal combustion engine” Graz, 2011. [76] Merker GP, Schwarz Ch, Teichmann R, Editors, Combustion engines development, 6th ed. Berlin, Heidelberg: Springer, 2012. doi: http://dx.doi. org/10.1007/978-3-642-14094-5. [77] Johannesson S, et al. ‘‘Gas Quality”, LNG Blue Corridors Project - European Comission - Seventh Framework Programme, GC.SST.2012.2-3 GA No. 321592, 2014, pp. 1–48. [78] CIMAC Working Group ‘‘Gas Engines”. Position Paper ‘‘Impact of Gas Quality on Gas Engine Performance”, Conseil International des Machines A Combustion, July 2015. [79] Cartellieri W, et al. ‘‘Erweiterung der Energieerzeugung durch Kraftgase – Teil 3”, Forschungsbericht Nr. 120, FVV Forschungsvereinigung Verbrennungskraftmaschinen, Frankfurt, 1971.

[80] Leiker M, et al. Evaluation of anti-knocking property of gaseous fuels by means of methane number and ist practical application to gas engines. ASME72-DGP-4, 1972. [81] Kubesh J, et al. Effect of gas composition on octane number of natural gas fuels. SAE Paper 922359, SAE International Fuels and Lubricants Meeting and Exposition, San Francisco, CA, USA, 1992. [82] Choquette G. Analysis and estimation of stoichiometric air-fuel ratio and methane number for natural gas. In: 23rd gas machinery conference, Nashville, TN, USA, 2014. [83] Andersen P. Algorithm for methane number determination for natural gasses. DGC project report, Danish Gas Technology Centre, Horsholm, DK, 1999. [84] Wimmer A, Chmela F, Kirsten M, Pirker G, Christiner P, Trapp C, et al. LEC-GPN - a new index for assessing the knock behavior of gaseous fuels for large engines. In: 4th international conference on knocking in gasoline engines. December 9 and 10, 2013, Berlin. [85] Kogler G, Wimmer A, Eichlseder H, Schnessl E, Winter H. Methodology in the development process of large gaseous fuelled engines. in: ICE 2003–0575 (2003), S. 547–557. Spring technical conference of the ASME internal combustion engine division, 2003. [86] Wimmer A, Pirker G, Schnessl E, Trapp C, Schaumberger H, Klinkner M. Bewertung von Simulationsmodellen zur Brennverfahrensauslegung für die neue Generation von Großgasmotoren. In: Presented at the 10th international symposium on combustion diagnostics, Baden-Baden, Germany, May 22–23, 2012. [87] Auer M, Friedrich C, Waldenmaier U, Knafl A, Stiesch G. Combustion development methodology for MAN diesel & turbo large bore gas engines. In: 14th conference ‘‘the working process of the internal combustion engine”, Graz 2013. [88] Schlemmer-Kelling U, Hamm T, Reichert E, Struckmeier D. FEV Single Cylinder Engine Family for large engine applications of MAN. In: 22nd Aachen colloquium automobile and engine technology 2013. [89] Trapp Ch. The future of gas engines – a technology comparison with diesel engines in the light of legislation and market. In: 14th conference ‘‘the working process of the internal combustion engine”, Graz, 2013. [90] Schlick H, Chmela F, Pirker G, Wimmer A. Weiterentwicklung eines nulldimensionalen Brennratenmodells für direktgezündete Gasmotoren auf Basis der Computertomographie, Tagung Motorische Verbrennung– München, 2009. [91] Chmela F, Krenn M, Pirker G, Schlick H, Wimmer A. Heat release rate simulation with non-natural gases. In: Proceedings, 8th Dessau gas engine conference, S. 197–210, 2013. [92] Rakopoulos CD, Michos CN. Development and validation of a multi-zone combustion model for performance and nitric oxide formation in syngas fueled spark ignition engine. Energy Convers Manage 2008;49:2924–38. [93] Wimmer A, Wohlthan M, Pirker G, Eichlseder H. Potential and limitations of cylinder pressure based engine testing and ways for plausibility checks. In: Presented at the 12th international symposium on combustion diagnostics, Baden-Baden, Germany; 2016 [94] Losonczi B, Chmela F, Pirker G, Wimmer A. Erkennung und Korrektur von Messfehlern zur Erhöhung der Genauigkeit von Motorprozessanalysen. In: 13th conference ‘‘the working process of the internal combustion engine”, Graz, 2011. [95] Wohlthan M, Pirker G, Sauperl I, Wimmer A, et al. Methodology for automated fault diagnosis at engine test beds. SAE Technical Paper 2017– 26-0326, 2017, doi:http://dx.doi.org/10.4271/2017-26-0326. [96] Wohlthan M, Pirker G, Wimmer A. Modular fault diagnosis system for engine test bed measurements. SAE Int. J. Commer. Veh. 2017; 10(1). doi:http:// dx.doi.org/10.4271/2017-01-0386. [97] Duclos J-M, Colin O. Arc and kernel tracking ignition model for 3d sparkignition engine calculations. In: S. I. Engine Combustion 7-Modeling, 5th Int. Symp. Diagn. Model. Combust. Intern. Combust. Engines COMODIA, 2001, 46. [98] Dahms RN, Drake MC, Fansler TD, Kuo T-W, Peters N. Understanding ignition processes in spray-guided gasoline engines using high-speed imaging and the extended spark-ignition model SparkCIMM. Part A: Spark channel processes and the turbulent flame front propagation. Combust Flame 2011;158:2229–44. [99] Lucchini T, et al. A comprehensive model to predict the initial stage of combustion in SI engines, SAE 2013-01-1087. [100] Schaefer L, Linse D, Durst B, Hasse C. 3D CFD simulation of the spark ignition process under the consideration of spark channel deflection, diffusion effects, curvature and detailed chemical kinetics. In: Int. Mot. 2016, 1st ed., Springer Vieweg, 2016. p. 784. [101] Meyer G, Gschirr A, Lindner-Silwester T, Stadlbauer K. Recent advances in modeling modulating capacitive high-energy ignition systems and application of the findings in a new generation of ignition systems. In: Eichlseder H, ediotr, 14th Conference ‘‘the working process of the internal combustion engine”, 2013. p. 393–408. [102] Meyer G, Gschirr A. Fundamental aspects and progress in initial flame kernel modeling for modern modulated capacity discharge high-energy ignition systems. In: Eichlseder H, editor, 15th Conference ‘‘the working process of the internal combustion engine”, 2015. p. 242–56. [103] Schneßl E, Pirker G, Wimmer A. Optimierung von Brennverfahren für Sondergasanwendungen auf Basis der Simulation und Versuch am Einzylinder-Forschungsmotor, MTZ-Konferenz Heavy Duty-, On- und OffHighway-Motoren, Friedrichshafen, 2009.

Please cite this article in press as: Pirker G, Wimmer A. Sustainable power generation with large gas engines. Energy Convers Manage (2017), http://dx.doi. org/10.1016/j.enconman.2017.06.023