LEAN BURN MILLER CYCLE GAS ENGINE COGENERATION SYSTEM Y. Matsushita, Osaka Gas Co., Ltd., Japan T. Fujiwaka, Osaka Gas Co., Ltd., Japan K. Tanaka, Mitsubishi Heavy Industries, Ltd., Japan T. Noguchi, Mitsubishi Heavy Industries, Ltd., Japan

1. INTRODUCTION

Generating efficiency　％

Recently, gas engine cogeneration systems which use town gas to supply heat and electrical power simultaneously have been introduced to industrial facilities that consume large amounts of thermal energy. In the past, the majority of customers operating cogeneration systems were those with a greater demand for heat than for electrical power. In recent years, however, technological progress in gas engines and gas turbines has led to high generation efficiency. Accordingly, as the advantages of cogeneration have come to be more widely recognized by customers with comparatively low demands for heat, its usage has become more common in a variety of fields. This trend is projected to accelerate in the future, and further improvement in the generation efficiency of gas engines and turbines is an urgent task. Against this background, Osaka Gas Co., Ltd. and Mitsubishi Heavy Industries, Ltd. have jointly applied the Miller cycle to a lean burn type gas engine, developing the world's first gas engine in this class to achieve a high generation efficiency. Figure 1 shows the generation efficiency of gas cogeneration systems in Japan and the target of the lean burn Miller cycle gas engine. The output range for gas engine cogeneration is mainly in the 100-1,000 kW class. At present, gas engines can be broadly classified as either the lean burn type1), the generation efficiency of which is 35 - 38 %, or the comparatively low-output stoichiometric type, the generation efficiency of which is 30 - 34 %. This paper reports the development of the new engine that features the world’s highest thermal efficiency of above 42 % in the under 1000kW class, and the performance specifications of the cogeneration system that attained the above described power generation efficiency, 40 % by using the new engine. 45

GT with steam on injecti injection

Target of Lean burn Miller cycle ＧＥ 40

35

Lean burn GE

30

Stoichiometric GE Simple cycle GT

25 100

200

500

1000

2000

5000

10000

Output　ｋＷ Figure 1. Generation efficiency of various cogeneration systems and the target of the new engine

2. MILLER CYCLE In conventional reciprocating internal combustion engines, where the compression ratio and the expansion ratio are approximately the same, the theoretical cycle efficiency is expressed as:

η=1-ε1-κ.

(1)

where ε is the engine expansion ratio (≒ compression ratio), and κ is the isentropic exponent of the working fluid. As shown in Formula(1), thermodynamically, the theoretical cycle efficiency η could be improved by increasing the expansion ratio ε. However, with conventional cycle engines, in which the compression ratio is equal to the expansion ratio, increase of the former follows that of the latter. Therefore, in gas engines which use natural gas as fuel, an excessively high expansion ratio ε that induces a high gas temperature in the cylinders causes knocking, whereby unburned gas spontaneously ignites during flame propagation. Rapid increases in cylinder temperature and pressure caused by knocking in turn cause damage to the engine. The Miller cycle2) was proposed as a means of resolving these difficulties. In the case of a conventional reciprocating internal combustion engine, the timing of closing the inlet valve and that of opening the exhaust valve are in the vicinity of the bottom dead center, and the compression ratio and expansion ratio are approximately the same. In the Miller cycle, as shown in Figure 2, the effective timing of closing the inlet valve is offset to slightly before or after bottom dead center, which serves to reduce only the actual compression ratio. This means that the expansion ratio is greater than the compression ratio. In other words, the application of the Miller cycle maintains the compression ratio at a level where knocking does not occur, and by promoting the larger expansion ratio, it enables high generation efficiency.

Conventional Cycle

Miller Cycle

Knocking e r u s s e r P r e n n I

× Compression stroke

Expansion stroke

Merit of Miller cycle e r u s s e r P r e n n I Volume

Intake valve closure timing

Compression stroke

Expansion stroke

Figure 2. Schematics of miller cycle and conventional cycle

Volume

3. DEVELOPMENT OF A HIGHLY EFFICIENT ENGINE There has been substantial research to date on the application of the Miller cycle3)-7). However, although some of these efforts have already led to practical applications, these cases have generally involved low thermal efficiency stoichiometric combustion type engines, and the practical application of thermal efficiency in excess of 40% has not been attained. In the development project reported here, the Miller cycle was applied to a pre-chamber type lean burn gas engine that features 37% engine output thermal efficiency even with a normal cycle, with the objective of obtaining a thermal efficiency of above 42 %. 3. 1. Test engine The test engine is a lean burn, six-cylinder gas engine with pre-chambers and a turbocharger (bore x stroke:170 mm x 180 mm) manufactured by Mitsubishi Heavy Industries, Ltd. Each cylinder has two intake valves and two exhaust valves, and a pre-chamber is attached at the center of the main chamber. An ignition plug is set into the sidewall of the pre-chamber. Natural gas is used for fuel. Fuel is directly fed into the pre-chamber, and a lean mixture is introduced into the main chamber. Combustion is initiated by the spark plug before the top dead center, and a combustion flame is then ejected as a torch jet from the pre-chamber to the main chamber. This system allows stable and efficient combustion of the lean mixture. To incorporate the Miller cycle concept, the following two improvements were implemented. First, to increase the expansion ratio, the volume of the piston cavity was reduced, thereby increasing the geometric compression ratio while maintaining a constant piston stroke. Secondly, to preset the expansion ratio and the compression ratio, the intake valve closure timing was set far from the bottom dead center. The specifications of the test engine and the properties of the fuel are listed in Table 1 and Table 2, respectively. Figure 3 show the combustion chamber of the test engine.

Cylinder Number

6

Bore x Stroke

170mm x 180mm

Mean Effective Pressure

1.2MPa

Engine Speed

1200min-1 / 1500min-1

Supercharging Method

Turbocharger

Table 1.

Engine specifications

Component

Volume (%)

CH4

88

C2H6

6

C3H8

3

C4H10

3

Table 2.

Properties of fuel component

Spark Plug Intake Valve

Exhaust Valve Pre-chamber

Piston

Main Chamber

Figure 3. Combustion chamber arrangement for the pre-chamber lean burn gas engine

3. 2. Optimization of the expansion and compression ratios Expansion and compression ratios were determined by means of an engine cycle simulator8) and the test engine. Figure 4 shows the relationship between the expansion ratio (based on the compression ratio at the bottom dead center) and thermal efficiency. Based on calculation results, we constructed several prototype pistons with expansion ratios ranging from 11 to 15 and verified them on a test engine. Calculations show that the thermal efficiency improves with increasing the expansion ratio. However, verification using the test engine indicates that knocking occurs at excessively high expansion ratios, resulting in decreases in the thermal efficiency. Accordingly, based on these findings, an expression ratio of 15 was selected. The timing of the inlet valve closure was then established experimentally. The relation between compression ratio (based on the compression at the inlet valve closure timing) and thermal efficiency was analyzed by using simulations and the test engine. An excessively high effective compression ratio is accompanied by knocking and requires delayed ignition timing, resulting in reduced thermal efficiency. On the other hand, an excessive reduction in the effective compression ratio results in a high boost pressure that leads to an increase in the pump loss and a reduction in thermal efficiency. On the basis of the results of these studies, the inlet valve closure timing was determined.

Thermal Efficiency (%)

43 42 41 40 39 38 37 36

Experiment Calculation Base

10

12 14 16 Expansion Ratio

18

Figure 4. Effect of expansion ratio on thermal efficiency

3. 3. Development of a high-efficiency turbocharger In order to reduce the effective compression ratio in Miller cycle gas engines, the opening area of the inlet valve is made smaller, and as a consequence the volumetric efficiency becomes lower. Thus, a higher inlet pressure is required in order to maintain the same output, and this necessitates a supercharger with a low flow volume and a high pressure ratio. However, no mass produced turbochargers were available with this combination of characteristics. When the turbocharger efficiency is low, regardless of how efficient the cycle within the piston is made, the exhaust pressure rises and the pump loss rate becomes large, such that no improvement in thermal efficiency can be realized. That is, a highly efficient turbocharger is a necessity for Miller cycle engines. A turbocharger optimized for Miller cycle use was accordingly developed. Optimization was carried out for each of the compositional elements of the turbocharger, namely, the turbine and the compressor, with the aim of achieving high efficiency. A new three-dimensionally optimized design was adopted for the blades of the turbine to reduce loss within the blades themselves and at the outlet, and efficiency was further boosted by using a wing shape for the nozzle and by optimizing the scroll shape. In the compressor, higher efficiency was also achieved by selecting a new three-dimensional shape for the impeller to reduce secondary flow loss, and by improving the scroll shape. The overall turbocharger efficiency of 62% was reached as a result, representing the world's highest level in this class. 3. 4. Improvement of combustion performance in the cylinder In order to increase the expansion ratio in Miller cycle gas engines, the volume of the combustion chamber at the top dead center is reduced. This promotes lower combustibility, thus canceling out the increased efficiency achieved with the higher expansion ratio. One of the key issues in developing a lean burn Miller cycle gas engine is how to ensure stable and fast combustion in the combustion chamber which has a small depth as a result of volume reduction. Stable and fast combustion is a prerequisite for high engine efficiency. In point of this view, the influence of the piston top configuration on the combustion was analyzed by using computational fluid dynamics and the test engine. Figure 5 shows some examples

of computational fluid dynamics in various shaped combustion chambers8). In pre-chamber type gas engines, spark ignition in the pre-chamber and the action of the torch flame are known to have a major influence on combustion in the main chamber. That is, the shape of the pre-chamber and the specifications of the blowhole substantially affect combustion efficiency and the heat release period. As illustrated in Figure 6, visualization technique of flames was utilized for the optimization of the pre-chamber shape (volume and blowhole). Based on the above-mentioned studies, a test for the optimization of combustion in a small volume space was conducted with respect to the details of the combustion chamber configuration, pre-chamber volume, nozzle angle, swirl intensity, and so forth. The combination of the technologies described above allowed the realization of an engine output efficiency of above 42% and a generation efficiency of 40%.

Figure 5. Calculated temperature profiles in the combustion chambers

Figure 6. Photographs of jet flames from pre-chamber

4. DEVELOPMENT OF A HIGHLY RELIABLE GAS ENGINE A high level of reliability is required for gas engines used in cogeneration systems because they are either operated continuously, or started and stopped every day. The following development efforts were carried out using a six-cylinder test engine to improve reliability. A cogeneration system with high reliability has been developed by adopting an electronic fuel control system and performing long-term endurance tests on its components. 4. 1. Electronic air-fuel control system In order to control the air-fuel ratio in the optimal range, it is necessary to control the respective flow volumes of the fuel gas and air. An electronic air-fuel control system was thus adopted for the Miller cycle gas engine developed in the project reported here. Based on the inlet manifold temperature, pressure, and engine speed, this system calculates the required supply volume of fuel gas, and controls this volume using a gas metering actuator. This allows control of the air-fuel ratio at an optimal level, regardless of changes in external air conditions, thereby ensuring high engine reliability and performance. Moreover, it also allows ease of start-up and loading characteristics, which are considered to be disadvantageous features of the lean burn engine. 4. 2. Development of high endurance spark plugs The newly developed engine has both a high combustion pressure and temperature compared with the prototype engine. Therefore, a high ignition voltage and a high temperature tolerance are required for the spark plugs. The conventional spark plugs had a lifetime of only 500 hrs in the newly-developed engine. Double-iridium spark plugs were developed and introduced into the market. Iridium chips are used in both the center electrode and the ground electrode. Iridium is a material not easily oxidized at high temperatures. The high-temperature tolerance characteristics increased substantially when the double iridium spark plug was adopted. Figure 7 shows the difference in ignition voltages between the conventional spark plugs used for a long time and the newly-developed spark plugs when they were tested in the engine developed in this study. The lifetime of the newly developed spark plugs is four times longer than that of conventional spark plugs. The newly developed spark plugs have a lifetime of over 2000 hrs. These spark plugs can be used not only in this engine but also in other types of engines.

40

Developed Plugs Conventional Plugs

Spark Voltage （ｋＶ）

35 30 25 20 15 10 0

500 1000 1500 2000 2500 Utilization period of spark plugs （hour）

Figure 7. Difference in endurance between conventional spark plugs and newly developed spark plugs.

4. 3. Endurance test An endurance test of longer than one year (10,000 hrs) was performed before introducing the system into the market. The newly developed engine maintains constant levels of electric-power-generation efficiency and heat-recovery efficiency throughout a one year period. We also conducted start-up tests during cold-temperature periods in the winter, and output-confirmation tests during an extremely hot summer. The newly developed engine can be started and operated even if outside weather conditions changed markedly. The engine was disassembled and examined after 10,000 hrs of operation. The piston, piston ring, connecting rod, intake valves, exhaust valves, pre-chamber valves, and other parts were examined in detail and there were no problems. Appropriate timing for the replacement of consumable items such as spark plugs, oil and filters was determined during the endurance test.

5. THE FEATURES OF THE COGENERATION SYSTEM USING THE LEAN BURN MILLER CYCLE GAS ENGINE Specifications of the newly developed lean burn Miller cycle gas engine and its package are listed in Table 3. This system has the following features. 5. 1. High power generating efficiency The lean burn Miller cycle gas engine has attained a thermal efficiency above 42%, which is 5 points above that of the conventional gas engine. Therefore, the cogeneration system of the new engine has the world’s highest power generation efficiency of over 40%, in the medium-scale class. 5. 2. Low heat to electricity ratio The total efficiency, including heat recovery, of the lean burn Miller cycle gas engine does not differ from that of the conventional lean burn engine. As a result, the heat to electricity ratio has decreased, since the power output is higher than the amount of heat recovered for the first time. Heat to electricity ratios of typical medium to small gas engines are shown in Figure 8. The heat to electricity ratio of the stoichiometric engine is as high as 1.7, as its generation efficiency is low. In contrast, the lean burn Miller cycle gas engine has a heat to electricity ratio of 0.86, about a half of that of the stoichiometric engine. It appears that this development trend will be accelerated in the future. The low heat to electricity ratio of the lean burn Miller cycle gas engine is expected to boost the use of cogeneration systems, as it helps to achieve the effective use of energy not only in the conventional cogeneration market but also in new areas such as office buildings and supermarkets, where the demand for heat is low and conventional cogeneration systems are not advantageous. Such new areas also include assembly plants and machining plants where the demand for electricity is high. 5. 3. Low impact to the environment Improvements have also been achieved in environmental friendliness. The increase in efficiency has achieved a reduction in CO2 emission of approximately 13% from that of conventional models, in terms of the amount of recovered heat. NOx concentration is also maintained at a low level of below 150ppm(O2=0%) using a simplified urea denitration developed by Osaka Gas. 5. 4. Compact system In response to the need for commercial cogeneration systems in office buildings, the cogeneration package is designed to consist of component units, and to be compact so that it is easy to carry into buildings in built-up areas. Wide spread use of the newly developed engine is anticipated due to its economical effectiveness and environmental friendliness.

Engine Name

GS6R, GS12R, GS16R

Type

4 stroke cycle

Cylinder dimension

L6, V12, V16

Comobustion type

Lean burn

Supercharging method

Turbocharger

Ignition type

Spark ignition

Bore(mm) x Stroke (mm)

170 x 180

Generating output (KW)

280-845

Heat recovery (KW)

241-759

-1

Engine Speed (min )

1200 / 1500

Expansion ratio (BDC besed)

15

Generating efficiency (%)

40

Heat recovery efficiency (%)

34.4

De NOx method

SCR using urea

NOx (O2=0%) (ppm)

150

Sound level (dB)

75

Table 3. Gas engine specifications

Generating output (kW)

1200

Lean burn Miller cycle GE

1000 800

Lean burn GE Stoichometric GE

600 400 200 0 0

200

400

600

800 1000 1200

Heat recovery (kW) Figure 8. Comparison of heat to electricity

6. FEALD PERFORMANCE

Generating efficiency (%)

Miller-cycle cogeneration systems with six different output levels ranging from 280 kW to 845 kW have been introduced to the market in 2001. Approximately 30 cogeneration systems were put into operation by February 2003 and they are reaching a total power-generation capacity of approximately 20,000 kW. An additional more than 30 systems are expected to be put into operation by the end of 2003. The first engine with an electric power output of 280 kW was installed on the roof of an office building. Currently, the system has been operating for longer than 6,000 hrs without any trouble. Since it is the first system, we examined changes in the ignition voltage at the plugs, changes in electric-power-generation efficiency, intake-exhaust valves, antechamber-valve wear and combustion stability. These results indicated that the system satisfied design-performance criteria in real commercial operations. Combustion was somewhat unstable when the engine was first introduced; this was corrected after modification of the electronic control system. At present, the average electric-power-generation efficiency of the 30 cogeneration systems operating in Japan is 40% or better. Figure 9 shows the changes in electric-power-generation efficiency in an engine installed at a certain site throughout a one-year period. The electric-power-generation efficiency did not change greatly regardless of the change of seasons; it stayed at 40% or greater. The newly developed system has a high electric-power-generation efficiency; these systems have already been introduced in various places. Three systems were installed in office buildings and four systems in supermarkets. These cogeneration systems are currently operating at sites where the introduction of these systems had been difficult because the locations required a greater supply of heat than electricity.

45 44 43 42 41 40 39 38 37 36 35 Winter

Spring

Summer

Autumn

Winter

Figure 9. Changes in electric power generation efficiency throughout a one year period

7. CONCLUSIONS Osaka Gas and Mitsubishi Heavy Industries have jointly applied the Miller cycle to a lean burn type gas engine, and have thus developed the world's first gas engine in this class to achieve a high generation efficiency. As a result of the development, the engine output efficiency reached above 42%. A gain of over 5 points was achieved in comparison with conventional lean burn engines, reaching the world's highest output efficiency level of 40%. The cogeneration system incorporating the new engine will have the following features. High power generation efficiency Low heat to electricity ratio Low impact to the environment Compact package With these features, this system will have high economic merit when used in office buildings which have comparatively small heat demands, where conventional systems previously failed to perform advantageously, as well as in the industrial sector where large and continuously increasing power demands are common. We are in the process of improving the output of electric power using currently-available Miller-cycle engines design as a starting point for development. Cogeneration systems which incorporate high-performance engines will be introduced to the market in 2003. In addition, we are developing high-efficiency engines with electric- power-generation efficiency exceeding 40%. In the future, the overall efficiency of gas engines is expected to improve. This encourages and increases the use of gas-cogeneration systems.

REFERENCES 1. Nakagawa, Mori, and Mizuta (1994). Studies on Pre-chamber Type Torch Ignition Lean Burn Gas Engines (The Impact of Pre-chamber Specifications on Combustion Properties) [in Japanese], Kikai Gakkai dai 72 ki zenkoku taikai koen-ronbun-shu Vol.3 p.356. 2. Miller, R.H. (1947). Supercharging and Internal Cooling Cycle for High Output, ASME 1947-6. 3. Sakai, H. et al. (1986). A Miller System Application for Efficient Diesel Power Units, IECEC, 1986 Vol.21 No.1. 4. Sakai, H. et al. (1985). A new type of Miller super charging system for high speed engines - Part 1 fundamental considerations and applications to gasoline engines, SAE851522. 5. Minoru Kamata et al. (1983). A new type of miller cycle gasoline engine. JSAE Review, July 1983, 16-21. 6. Tsuki et al. (1993). Development of a V6 Miller Cycle Engine [in Japanese], Jidosha Gijutsu Kai Gakujutsu koen-kai zen-sasshu 935 1993-10. 7. FuRong Zhang, K.Okamoto, S.Shimogata, F.Shoji (1996). Development of The Miller Cycle for Natural Gas Engines, ICE-Vol.27-1, 1996 Fall Technical Conference Volume 1 ASME. 8. Fujiwaka et al. (1998). Development of the Miller cycle lean burn gas engine, IGRC1998.