Printer Friendly

The effect of cooled exhaust gas recirculation for a naturally aspirated stationary gas engine.

ABSTRACT

Small natural gas cogeneration engines frequently operate with lean mixture and late ignition timing to comply with [NO.sub.x] emission standards. Late combustion phasing is the consequence, leading to significant losses in engine efficiency. When substituting a part of the excess air with exhaust gas, heat capacity increases, thus reducing [NO.sub.x] emissions. Combustion phasing can be advanced, resulting in a thermodynamically more favourable heat release without increasing [NO.sub.x] but improving engine efficiency.

In this work, the effect of replacing a part of excess air with exhaust gas was investigated first in a constant volume combustion chamber. It enabled to analyse the influence of the exhaust gas under motionless initial conditions for several relative air-fuel ratios ([lambda] = 1.3 to 1.7). Starting from the initial value of [lambda], the amount of [CH.sub.4] was maintained constant as a part of the excess air was replaced by exhaust gas. Temperature of burnt and unburnt mixture, combustion velocity and others were analysed for varying EGR rates.

In the next step, cooled exhaust gas recirculation was employed for a single-cylinder cogeneration engine to evaluate the influence of EGR rates up to 25 % on engine performance and emissions. For the baseline compression ratio of 13.2, a slightly higher IMEP could be achieved coupled with lower ISFC (almost -5 g/kWh) at constant [NO.sub.x] emissions. Subsequently, the compression ratio was raised to 15.25 and the trials repeated, showing further improvement with the drawback of higher knock tendency.

CITATION: Neher, D., Scholl, F., Kettner, M., Schwarz, D. et al., "The Effect of Cooled Exhaust Gas Recirculation for a Naturally Aspirated Stationary Gas Engine,"

INTRODUCTION

The simultaneous use of heat and power allows cogeneration units (or also known as combined heat and power units) to achieve overall efficiencies of over 90 %. Therefore, it represents an attractive approach to reduce greenhouse gas emissions according to the defined target of the European Union (EU). The aim of 80-95 % C[O.sub.2] reduction by 2050 relative to 1990 levels was reinforced by publishing the low-carbon roadmap in 2011 [1]. In Denmark, Finland, Russia and Latvia 30-50 % of the generated power was achieved through cogeneration units in 2008. A substantial growth is also to be expected, among others, in USA, Japan, Russia and Brazil [2].

The systems are mostly equipped with natural gas (NG) engines, providing high thermal efficiency and long-term durability. The emission legislations differ worldwide significantly (even within certain countries such as Switzerland or Japan), depending on generated electrical power or thermal energy input. In Germany, cogeneration manufacturers of the small power range orient themselves to emission standards according to TA-Luft [3]. It obliges large gas engines with a thermal input of 1 MW to meet dry [NO.sub.x] emissions as low as 500 mg/[mN.sup.3] at 5 % [O.sub.2] (approx. 200 ppm). The significant rise of installed systems over the past years leads to progressively stringent emissions standards. In 2018, a Europe-wide tightening of emissions standards will take place, applying for NG driven engines with a generated electrical power of [P.sub.el] = 50 kW and less [4]. The [NO.sub.x] limit of 240 mg/kWh dry at 0 % [O.sub.2] is related to thermal input (higher heating value HHV), amounting to roughly half of the present value according to TA-Luft (i.e. approx. 100 ppm).

For the small power sector of 50 k[W.sub.el] and less, naturally aspirated gas engines conventionally operate unthrottled with external mixture formation with either lean or stoichiometric air-fuel mixture. The latter requires a three-way-catalyst (TWC) in order to reduce tail pipe emissions such as hydrocarbons (HC), carbon monoxide (CO) and nitrogen oxide ([NO.sub.x]). In lean burn operation, on the other hand, HC and CO emissions are converted through an oxidation catalyst, while low [NO.sub.x] emissions result from low combustion temperature. This is not only ensured through a lean mixture but also by retarded ignition timing, leading to a thermodynamically unfavourable centre of combustion (CA50). Since combustion temperature and thus heat wall losses drop with leaning, the engine still reaches relatively high thermal efficiency, in spite of operation with relative air-fuel ratios [lambda] beyond the efficiency optimum.

Figure 1 shows the trade-off between engine efficiency, [NO.sub.x] emissions and IMEP for a naturally aspirated engine. Further mixture dilution allows the engine to meet future emission standards, however, combustion progressively deviates from ideal Otto cycle and cycle-by-cycle variations increase, leading to lower thermal efficiency. Leaning also reduces fuel fraction for naturally aspirated engines, resulting in lower IMEP and therefore in a higher impact of friction, reducing engine's brake efficiency.

This paper provides an analysis of an alternative mixture configuration for small-scaled cogeneration engines with the aim of improving the previously described trade-off. Exhaust gas is recirculated into the intake section to replace merely a part of the excess air, increasing mixture's heat capacity. As a result, engine efficiency can be improved by advancing combustion phasing at equal [NO.sub.x] emissions.

The paper begins with the fundamentals of exhaust gas recirculation, addressing the definition of lean burn and revealing the influence of EGR on the relative air-fuel ratio. The results of 0D engine calculations suggest the theoretical potential of EGR for lean burn operation. Subsequently, experiments of different mixtures compositions using a constant volume combustion chamber are presented in order to show the effect of mixture's composition on combustion, in particular on its burning velocity, under motionless conditions. The next chapter covers the experimental setup used for engine trials, followed by experimental results for the baseline but also for an increased compression ratio. Eventually, a brief summary is given.

EXHAUST GAS RECIRCULATION

EGR represents an established means to dilute air-fuel mixture. It is widely used whether for Otto engines to lower pumping losses in part load operation or for Diesel engines to reduce combustion temperature and thus to counter [NO.sub.x] emissions. The formation of [NO.sub.x] is inhibited as a result of cooling the combustion through so-called mass and caloric effect [5, 6]. The first describes the increasing stoichiometric mass of the non-premixed mixture that combusts at approximately [lambda] = 1, while the caloric effect arises from the higher specific heat capacity due to the high [H.sub.2]O and C[O.sub.2] amount of the exhaust gas [5, 6] (see Figure 2).

In contrary to Diesel, natural gas engines emit practically no soot [7] that could contaminate components such as EGR valve or cooler and appears therefore to be suitable for long-term operation.

Work of Others

Numerous comparisons between both operation modes pure lean burn and stoichiometric with EGR can be found in literature [8, 9, 10]. In contrast to the mentioned work, the subject of this paper is to combine lean burn operation and EGR as (only) a part of the excess air is substituted with exhaust gas. To date, this approach was primarily investigated in the course of research projects [5, 6, 7, 10, 11, 12, 13, 14, 15, 16]. Numerical work points out potential in reducing [NO.sub.x] emissions, while efficiency behaviour depends on engine size and the type of EGR strategy. Lee et al. [11] investigated a dedicated EGR system that reduces fuel consumption of small 4-6 cylinder engines with 338 [cm.sup.3]/cylinder, however, the dedicated EGR cylinder was operated with rich air-fuel mixture while the other cylinders operated with lean ([lambda] < 1.3) or stoichiometric mixture, respectively. Wimmer [5] and Pizzirani et al. [12] studied the effect of EGR for large lean-burn gas engines, concluding that [NO.sub.x] emissions significantly fall at the slight expense of engine efficiency. Since knocking tendency also decreases, Pizzirani et al. see potential in increasing efficiency by augmenting compression ratio.

Wimmer also studied the effect of EGR experimentally for a single-cylinder research engine (6.2 litres) in stoichiometric and lean burn operation [6]. Lean burn operation coupled with moderate EGR rates (15 %) reduces [NO.sub.x] emissions notably while efficiency suffers only slightly due to imperfect combustion and altered charge composition. Tschalamoff [13, 13, 14] employed EGR (EGR rates up to 30 %) for a single-cylinder 4.83 litre engine with scavenged and unscavenged prechamber spark plug. [NO.sub.x] emissions and mean effective pressure (MEP) were maintained constant by adjusting fuel quantity and ignition timing, respectively. For an unscavenged spark plug, engine efficiency remains unchanged, whereas it increases in the case of scavenged prechamber spark plug by 1.5 %-points. The benefit was found in thermodynamically more favourable CA50 at same [NO.sub.x] emissions affected by higher specific heat capacity of the mixture. Since temperature is also one of the driving parameters for knocking combustion, EGR extends the knock margin for both lean [5, 13, 14, 17] and stoichiometric operation [6, 7, 9, 17]. This allows the engine to operate at higher compression ratio and/or higher IMEP.

Definition of Mixture Dilution and EGR Rate

For the later course of this paper, the parameters to describe mixture composition are of vital importance. The amount of exhaust gas is expressed by the EGR rate [x.sub.E][G.sub.R], determined by dividing the volume flow of the recirculated exhaust gas [V.sub.EGR] with the aspirated mixture, consisting of the volume flow of exhaust gas [V.sub.EGR], air [V.sub.air] and fuel [V.sub.fuel]:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

The relative air-fuel ratio [lambda] or its inverse, the fuel-air equivalence ratio [phi], are commonly used parameters for defining mixture composition. These are formulated by the ratio of moles of oxygen ([O.sub.2]) of the in-cylinder charge [n.sub.O2] and the required moles of [O.sub.2] for a stoichiometric combustion [n.sub.O2 stoich] or vice versa:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (2)

Stoichiometric combustion takes place at [lambda] = 1.0, while [lambda] > 1.0 is defined as lean combustion. When substituting a part of excess air of a lean mixture with exhaust gas, the amount of fuel remains unchanged, leading to a lower relative air-fuel ratio thanks to the smaller [O.sub.2] fraction of the exhaust gas compared with air. For better understanding of the dependency between EGR rate and relative air-fuel ratio, the dilution ratio D is introduced

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)

where [n.sub.air] and [n.sub.EGR] are the number of moles of air and EGR of the in-cylinder charge and [n.sub.air stoich] is the number of moles of air that results in a stoichiometric combustion.

Figure 3 (left) shows the influence of D and [x.sub.EGR] on the relative air-fuel ratio quantitatively, while Figure 3 (right) illustrates a mixture with a dilution ratio of D = 1.40 over an EGR rate from 0 to 20 % schematically. In the case of pure mixture dilution through air, the relative air-fuel ratio matches with the dilution ratio [lambda] = D = 1.40. With rising EGR fraction but constant dilution ratio D = 1.4, the amount of O2 decreases, leading to lower air -fuel ratios of [lambda] = 1.28 and [lambda] = 1.12 for an EGR rate of 10 and 20 %, respectively.

0D ENGINE CALCULATION

For the ideal engine cycle, compression ratio and mixture's composition determine engine efficiency. Therefore, the effect of dilution ratio and EGR rate on the trade-off between engine efficiency, IMEP and [NO.sub.x] emissions was studied using a 0D engine calculation model (created with AVL BOOST). It considers the cylinder as adiabatic system with perfect in-cylinder filling (i.e. no residual gas) and assumes isochoric combustion and gas exchange. The computation considers the temperature-driven change of thermochemical properties, while the initial temperature and pressure of the air-methane-exhaust gas mixture are defined to [T.sub.0] = 293.15 K and [p.sub.0] = 1.0 bar. The water of the exhaust gas is assumed to have a gaseous phase at initial conditions, which is only a minor error due to the rapid temperature rise as compression takes place.

Figure 4 shows thermal efficiency, peak combustion temperature and normalised indicated work over compression ratio (CR = 10 to 18) for several mixture settings. Due to the high temperature dependency of [NO.sub.x] emissions, peak combustion temperature represents their indicator. As expected, peak combustion temperature, indicated work and thermal efficiency increase with compression ratio. An engine setup characterised by a dilution ratio of D = [lambda] = 1.50 without EGR and a compression ratio of CR = 13.0 is a commonly used setting for a series operating point of small gas engines, thus serving as a reference for comparison. When replacing a part of the excess air of the reference configuration with exhaust gas (D = 1.50/[x.sub.EGR] =20 %), peak combustion temperature decreases in spite of same quantity of fuel added to the cycle. This arises from the higher specific heat capacity of the in-cylinder charge, absorbing a larger amount of heat as combustion takes place. As a result, peak combustion pressure drops, reducing cycle work and efficiency. A rise of the compression ratio (CR = 13. 50) compensates losses in efficiency, while indicated work is still below the reference operating point. This is contributable to the smaller amount of fuel added to the cycle as compression ratio increases and therefore cylinder volume at BDC decreases. The cycle outputs the same amount of work when further enhancing compression ratio up to CR = 13.95. Engine efficiency increases by 0.4 %-points while peak combustion temperature remains significantly lower.

A higher dilution ratio of D = [lambda] = 1.60 (i.e. [x.sub.EGR] = 0%) necessitates a compression ratio of 17.80 to achieve equivalent indicated work as the reference due to lower amount of fuel. Peak combustion temperature lies slightly above the one of D = 1.50/[x.sub.EGR] = 20 % at CR = 13.95, whereas the benefit in efficiency is significantly higher. A lower dilution ratio (D = 1.40) without EGR results in peak combustion temperature higher than the value of the reference operating point. It decreases significantly by adding EGR ([x.sub.EGR] = 20 %), however, the reference temperature is only achieved by also reducing the compression ratio and thus deteriorating thermal efficiency notably.

In both cases D = 1.60/[x.sub.EGR] = 0 % and D = 1.50/[x.sub.EGR] = 20 % high compression ratios (CR = 17.8 and 13.95 respectively) are calculated to maintain indicated work of the reference operating point. Due to the high knock resistance of natural gas as well as the knock-reducing effect of EGR known from literature [5, 13, 14, 17], a compression ratio of CR = 13.95 appears to be a feasible value with respect to real engine operation (without knocking). A configuration of leaner mixture and even higher compression ratio (D = 1.60/[x.sub.EGR] = 0 % at CR = 17.80), on the other hand, results in similar combustion temperatures but would increase the chance of knocking in real engine operation significantly. This arises from its dependence on pressure and temperature history as well as on combustion speed, which however were neglected in this study.

EXPERIMENTS WITH A CONSTANT VOLUME COMBUSTION CHAMBER

The laminar burning velocity of a premixed air-fuel-exhaust gas mixture highly influences the heat release during the cycle and therefore engine efficiency and its emissions. It only depends on composition, temperature and pressure of the mixture. It is defined as the velocity, with which unburnt gas of the premixed mixture propagates through the flame front and is transformed to products under laminar flow conditions [18, 19]. To determine the laminar burning velocity of various mixture compositions, experiments using a constant volume combustion chamber (CVCC) were carried out at the University of Valladolid, Spain.

Experimental Setup and Numerical Analysis

The CVCC is made of stainless steel and has a spherical shape with a inner diameter of 200 mm, corresponding to a 4.19 litre volume. It is connected to several gas bottles as well as an air compressor on the inlet and a vacuum pump on the outlet side (see Figure 5). To measure the initial conditions of pressure and temperature, four piezoresistive sensors are used and a thermocouple is located in the CVCC. The CVCC is thoroughly insulated and equipped with an electrical heater on the bottom, enabling initial mixture temperatures up to 650 K. Two prolonged electrodes form the spark gap in the centre of the CVCC, leading to a uniform flame propagation over the vessels radius. The pressure signal over time is measured by the means of a Kistler 7063A pressure transducer that is connected to an oscilloscope for recording the data.

The data is evaluated using a proprietary two-zone combustion analysis model, dividing the gas within the CVCC into burnt and unburnt fraction. The model considers exhaust gas of the initial mixture as products of perfect combustion C[O.sub.2], [N.sub.2], [O.sub.2] and [H.sub.2]O, whereas for combustion itself it solves 12 equilibrium equations of the major species. The thermochemical properties taken from NIST-JANAF tables consider temperature dependency of the specific heat capacity. The calculation model assumes following behaviour:

* Combustion commences in the centre of the vessel, leading to a flame front that propagates spherically with negligible thickness.

* The pressure during combustion is uniform throughout the whole vessel.

* The CVCC is an adiabatic system, meaning that no heat is lost through the walls.

* Ideal gas behaviour is assumed.

* Unburnt gas is isentropically compressed as combustion proceeds.

The first law of thermodynamics for a non-stationary open system is applied to burnt and unburnt zone, describing their change of internal energy. After adding mass conservation and volume equation, the laminar burning velocity Cc yields to Eq.(4)

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (4)

where the mass burning rate [dm.sub.b]/dt is achieved from pressure data, the flame area [A.sub.f] is determined over the burnt volume by assuming spherical flame shape and [[rho].sub.u] is the density of the unburnt zone. For the derivation of Eq.(4) as well as a more detailed description of the pressure analysis model and the experimental setup, the authors refer to former work [18, 20, 21, 22]

Testing Procedure

All experiments were conducted at same initial temperature of [T.sub.0] = 423.15 K and initial pressure of [p.sub.0] = 3.73 bar, resulting in a constant amount of moles. The initial conditions were determined by isentropic change of state starting from 293.15 K and 1.0 bar. The dilution ratio was varied from D = 1.3 to D = 1.8 in increments of 0.1. For each dilution ratio, a part of the excess air was substituted with exhaust gas, starting from an EGR rate of [x.sub.EGR] = 5 % up to [x.sub.E][G.sub.R] = 20 %.

After every trial, the CCVC was scavenged for a duration of 5 minutes using compressed air to avoid any remaining combustion products. The vessel was then evacuated and refilled first with CH4 and after with compressed air. The corresponding partial pressures of each mixture were calculated by assuming ideal gas behaviour. After a waiting time of 5 minutes, the mixture was ignited. Prior investigations proved this to be a sufficient duration to achieve well-mixed and motionless mixture [18, 19].

Experiments including EGR necessitated an additional working step. In order to obtain exhaust gas components such as [H.sub.2]O and C[O.sub.2] within the vessel, first a stoichiometric combustion was conducted. After a waiting time of 3 minutes the desired partial pressure to obtain the required C[O.sub.2] and [H.sub.2]O fraction from the former combustion was adjusted using the vacuum pump. Subsequently, [CH.sub.4] and air were added, including the missing excess part of N2 and [O.sub.2] of the exhaust gas after the stoichiometric combustion.

Results of Varying Mixture Composition

Figure 6 shows pressure versus time for all investigated dilution ratios without exhaust gas (left) (D = [lambda]) and for a dilution ratio of D = 1.4 of varying EGR rates (right) at constant ignition timing [t.sub.ignition] = 0.00 s. Under the premise of constant initial conditions, the amount of fuel falls for higher dilution ratios. Peak and gradient of the combustion pressure decrease, leading to retarded peak combustion pressure. A similar behaviour is observed when exhaust gas is added for constant dilution ratio. Specific heat capacity increases for higher EGR rates, resulting in lower pressure rise with a lower pressure peak and retarded combustion.

Figure 7 shows the calculated burning velocity Cc against the temperature of the unburnt zone [T.sub.u] of selected mixture compositions. Note that same [T.sub.u] corresponds to very similar pressure thanks to the isentropic calculation of [T.sub.u]. The first part of the plotted traces between the initial temperature and approximately 475 K is futile with respect to evaluate mixture's burning velocity due to instability of the values. Burning velocity decreases with higher dilution ratio and higher EGR rate significantly. The lower peak in combustion pressure as a result of leaning or adding EGR, respectively, correlates with the maximum temperature of the unburnt zone, thus narrowing the evaluation window of burning velocity. Therefore, only dilution ratios of D = 1.3 to 1.5 with and D = 1.6 without EGR are plotted in Figure 7. After burning velocity reaches its maximum, it declines rapidly until combustion is completed. The gradient of this section decreases with both higher dilution ratio and EGR rate, enhancing the time that the flame front dwells in the vicinity of the wall and thus promoting the chance of flame quenching.

Figure 8 (left) depicts mixture's initial specific heat capacity [c.sub.v] (before ignition) against the EGR rate for various dilution ratios D. For a constant D, i.e. constant mass of fuel, the specific heat capacity increases for higher EGR rates. It rises stronger for lower dilution ratios due to the higher C[O.sub.2] and [H.sub.2]O amount of the mixture. When dilution ratio increases for same EGR rate, [c.sub.v] of the mixture becomes smaller caused by lower [CH.sub.4] fraction (cf. Figure 2). This, in turn, leads to a larger excess fraction that absorbs the heat during combustion, resulting in longer burning delay (time until 5 % of the mass fraction is burnt) as shown in Figure 8 (right).

The burning delay increases with higher dilution ratio and EGR rate, as shown in Figure 8 (right). The difference between dilution ratio at same EGR rate decreases for higher exhaust gas fraction, since exhaust gas of richer mixture reveals higher specific heat capacity.

Figure 9 depicts burning velocity as function of the EGR rate and the temperature of the burnt zone [T.sub.b] for various dilution ratios at constant temperature of the unburnt zone of [T.sub.u] = 525 K. As expected, the lowest dilution ratio D = 1.3 reaches the peak values in temperature of the burnt zone of [T.sub.b] = 2250 K and combustion velocity of Cc = 0.325 m/s. For a dilution ratio of D = 1.6 and no EGR for instance, both values drop down to [T.sub.b] = 2044 K and Cc = 0.132 m/s, respectively. By replacing air with exhaust gas, EGR enables to cool the combustion while burning velocity decreases significantly.

The required EGR rate for same temperature of the burnt zone increases for lower dilution ratios. The peak in burning velocity of Cc = 0.255 m/s at [T.sub.b] = 2175 K is found for a dilution ratio of D = 1.4 and an EGR rate of approximately [x.sub.EGR] = 3 %. When further lowering the dilution ratio down to D = 1.3, the required EGR rate rises to almost [x.sub.EGR] = 19 %, reducing burning velocity notably down to Cc = 0.191 m/s.

It is well-known that the shape of the flame front may change in the course of the combustion due to instabilities caused by hydrodynamic effects [23, 24]. These lead to flame cracking, which is a precursor to a cellular flame structure. The latter exhibits a larger surface than the purely laminar and spherical flame shape, accelerating the combustion. In that case, the analysis code calculates a higher value than the actual one, since it still assumes a laminar flame front (cf. Eq.(4)). Nevertheless, preliminary investigations at D = 1.4 to 1.6 showed that cellularity is unlikely to occur.

The experiments point out that the burnt zone temperature of lean mixtures can be achieved also for mixtures of higher fuel fraction by replacing a part of excess air with EGR. As a result, burning velocity decreases, which might be compensated by advancing ignition timing in engine operation. In that case, the crucial differences are higher temperature and pressure level as well as motion of in-cylinder charge, leading to turbulent combustion.

TEST BED ENGINE

The potential of EGR was studied at Karlsruhe University of Applied Sciences using the series production cogeneration engine SenerTec Dachs 5.5, which was equipped with metrology and an EGR system.

Engine Specifications and Metrology

The naturally aspirated single-cylinder engine operates permanently unthrottled at a constant engine speed of n = 2450 rev/min, generating 5.5 kW of electrical power (equivalent to IMEP = 6.35 bar) in its reference operating point (ROP). At a relative air-fuel ratio of [lambda] [approximately equal to] 1.50 and a fixed ignition timing IT = -8 [degrees]CA the centre of combustion (defined by 50 % mass fraction burnt) occurs at CA50 [approximately equal to] 19 [degrees]CA after top dead centre ATDC, leading to [NO.sub.x] emissions of approximately 1.44 g/kWh. Note that these settings will serve as a reference when comparing different engine configuration in the later course of this paper.

The engine has a heron design cylinder head and a compression ratio of CR = 13.20. In order to study the effect of EGR also for a higher compression ratio of CR = 15.25, a prototype piston was developed by reducing piston bowl's depth and slightly widening its diameter. Results of prior investigations showed that the area of the piston bowl [A.sub.pb] is the decisive parameter with respect to [NO.sub.x] emissions. [A.sub.pb] was kept constant at the expense of squish ratio [C.sub.sq] and surface area to volume ratio A/V. In order to ensure stable and fast combustion, an unscavenged prechamber spark plug ignites the lean air-NG mixture.

The series production engine is the baseline of this work. Its technical specifications are listed in Table 1.

The in-cylinder pressure and intake port pressure over crank angle were measured using Kistler pressure transducers of type 6125C11 and 4007B, both connected to an AVL Indimodul 622 indicating system. It calculates all relevant combustion parameters based on the first law of thermodynamics for a predefined number of 200 cycles. Knocking detection was conducted by analysing the cylinder pressure signal. The signal of every cycle is high pass filtered with a cutoff frequency of 4 kHz and rectified to determine the peak value of the superimposed oscillations. If the value exceeds the predefined threshold of 0.25 bar, knocking combustion is detected. The air mass flow was monitored using a hot-film air-mass meter model HFM 5 from BOSCH calibrated to measure small mass flows. A Bronkhorst F-103 thermal mass flow meter for NG was used for measuring the fuel consumption.

Analysis of engine-out emissions were measured with an exhaust gas analyser ABB AO2000, which determines the concentrations of C[O.sub.2], CO and [O.sub.2] under dry and of NO, N[O.sub.2] and [C.sub.3][H.sub.8] under wet conditions. The knowledge of the single exhaust gas components enables to calculate the relative air-fuel ratio quite accurately and to analyse the exhaust gas that is fed back into the intake section. A conditioning unit allows the engine to operate at constant inlet conditions of the cogeneration unit (CU) (see Figure 10). The natural gas composition of the data presented in this paper was monitored using an Inficon Micro GC Fusion Gas Analyser. During trials, only minor variations in gas quality were registered, as shown in Table 2.

Engine's Gas Path, EGR System and EGR Rate Model

The engine is encapsulated to ensure low noise emissions, leading to relative high casing temperatures of approximately [T.sub.casing] = 85 [degrees]C. The air-NG mixture prepared in a venturi mixer outside of the casing heats up as it flows through the relatively long gas path of more than 2 m. It exhibits a mean temperature of [T.sub.port in] = 74-75 [degrees]C in the intake port, shortly before the intake valve. After combustion, the burnt mixture escapes from the cylinder and flows into an oxidation catalyst, followed by a heat exchanger and a muffler.

The sampling of the exhaust gas analyser was placed between exhaust port and oxidation catalyst. To ensure pure engine-out emissions are monitored, the oxidation catalyst was replaced by a dummy component of same geometry but without catalytic coating.

To investigate the impact of exhaust gas on the generally purely lean operating baseline engine, an EGR system needed to be incorporated. The exhaust gas is extracted at the outlet of the heat exchanger, where its temperature amounts to 160 [degrees]C and the pressure level is still high enough to enable EGR rates up to 25 %. The temperature of the recirculated exhaust gas depends on coolant temperature of the EGR cooler, designed as shell and tube heat exchanger. A process thermostat allows the coolant's temperature to be varied over a wide range. Its minimum temperature is limited to 58 [degrees]C to prevent condensation of the exhaust gas within the EGR cooler.

The amount of exhaust gas that is fed back into the intake system just before the intake port is regulated through an EGR valve. The EGR rate is often determined by the ratio of C[O.sub.2] concentration of the air-fuel-exhaust gas mixture and the exhaust gas [9, 16]. This approach, however, requires a homogenous air-NG-exhaust gas mixture and is because of the short mixing section combined with small installation space not applicable. Therefore, the EGR rate (defined by Eq.(1)) is determined using a calculation model that assumes a constant reference volume flow of the mixture [V.sub.ref] under the premise of constant engine speed und unthrottled engine operation. The reference volume flow is first calculated for engine operation without EGR by the sum of [V.sub.air] and [V.sub.fue]. Both are calculated by the measured values of air and fuel mass flow as well as their temperature and static pressure. The change in volume flow when adding exhaust gas to the intake is entirely caused by EGR itself, following to:

[[??].sub.EGR] = [[??].sub.ref] - [[??].sub.fuel] -[[??].sub.air] (5)

The reference volume flow was evaluated for several operating points of varying CA50, air-fuel ratio and IMEP, showing only small deviations. Besides, detailed 1D-CFD calculations proved this a legitimate approach for unthrottled operation at constant engine speed, leading to a maximum error lower than 0.5 %-points for an EGR rate of [x.sub.EGR] = 25 %.

The fact that the air-fuel-exhaust gas mixture homogenises only over a short mixing section before it enters the combustion chamber, encouraged to conduct 3D-CFD simulations. These indicate that as the mixture flows through the swirl port into the cylinder, a pronounced motion takes place, thus ensuring sufficient homogenisation.

ENGINE RESULTS

All engine trials were carried out under full load operation at constant engine speed of 2450 rev/min. Pressure and temperature of the cogeneration unit were maintained constant at [p.sub.CU in] = 995 mbar and temperature of [T.sub.CU in] = 35 [degrees]C, while ambient pressure, the exhaust system boundary of the engine, varied only slightly.

In order to obtain high validity of the experimental results, every operating point was measured three times over a period of 30 seconds. Tail pipe emissions are shown in g/kWh (related to indicated power), while HC emissions show methane hydrocarbons.

The target of the investigations was to analyse the effect of EGR on the trade-off between engine efficiency, [NO.sub.x] emissions and IMEP. Since a loss in IMEP increases the ratio between investment costs and electrical power output of the cogeneration unit, its product attractiveness would suffer. Therefore, the primary focus was placed on engine trials of constant IMEP = 6.35 bar (corresponding to the reference operating point ROP) with respect to a potential series introduction of EGR. However, to ensure that the whole potential of EGR is found regardless of IMEP, the latter was varied for constant [NO.sub.x] emissions.

The EGR rate was adjusted from 0 to 25 % stepwise in increments of [DELTA][x.sub.EGR] = 5 %. In order to avoid a significant influence of EGR on intake port temperature [T.sub.port in] and thereby to isolate the effect of EGR on engine characteristics, the exhaust gas was recirculated at [T.sub.EGR] = 75 [degrees]C. Only in the case of EGR rates above [x.sub.EGR] = 20 %, the required coolant temperature would fall below its lower limit of 58 [degrees]C, causing condensation of the exhaust gas. The highest investigated EGR rate of [x.sub.EGR] = 25 % resulted in [T.sub.EGR] = 78 [degrees]C, affecting intake port temperature only slightly. The minor temperature difference between EGR and air-NG mixture also leads to similar density and therefore the volumetric EGR rate differs marginally to the mass related EGR rate.

Baseline Piston - CR = 13.20

This chapter covers the results of the baseline piston of the series production engine including EGR rates up to 20 %.

Constant IMEP

For each EGR rate, CA50 sweeps were conducted while IMEP was maintained constant by adjusting ignition timing and dilution ratio (i.e. mass of fuel affected). CA50 between 6 and 10 [degrees]CA ATDC represents the high efficiency regime, at which the engine operates at its highest [NO.sub.x] emissions. Retarded CA50 of 16 to 20 [degrees]CA ATDC, on the other hand, lead to low [NO.sub.x] emissions at the expense of efficiency. Both low [NO.sub.x] and high efficiency regime are of high interest in this study and were therefore recorded in steps of [DELTA]CA50 = 1 [degrees]CA, whereas in-between only CA50 = 13 [degrees]CA ATDC was taken into account.

Figure 11 depicts ISFC (top) and CA50 (bottom) over [NO.sub.x] emissions for different EGR rates at constant IMEP = 6.35 bar (+/- 0.01 bar). With increasing EGR rate, the Pareto front further shifts toward lower [NO.sub.x] emissions and higher ISFC, as shown in Figure 11 (top).

Figure 11 (bottom) shows that the increasing mixture's specific heat capacity caused by EGR decreases [NO.sub.x] emissions at same CA50. To maintain engine's efficiency, the required CA50 advances for higher EGR rates while [NO.sub.x] emissions still diminish. Even for earlier combustion phasing, [NO.sub.x] emissions are lower than for the ROP, favouring engine's fuel consumption. The largest improvement in [NO.sub.x] at same efficiency is found for EGR rates of 15 and 20 %, requiring CA50 of 16 [degrees]CA and 18 [degrees]CA ATDC respectively. Compared with the reference operating point ([NO.sub.x] = 1.44 g/kWh, ISFC = 206.7 g/kWh) ISFC and [NO.sub.x] emissions result to 204.1 g/kWh and 1.28 g/kWh for [x.sub.EGR] = 15 % and to 207.3 g/kWh and 1.07 g/kWh for [x.sub.EGR] = 20 %.

The impact of CA50 on engine efficiency becomes smaller for the high efficiency regime. Regardless of the EGR rate, the lowest ISFC is achieved at CA50 = 8 [degrees]CA ATDC (see Figure 11 (bottom)). The influence of CA50 on ISFC is rather small whether CA50 is 2 [degrees]CA advanced to or retarded from the efficiency optimum. [NO.sub.x] emissions, on the other hand, increase notably, deteriorating the trade-off within the high efficiency regime as EGR rate increases.

To ensure that the altered engine characteristics caused by EGR majorly result from altered mixture composition rather than lower pumping losses, the gas exchange was analysed. When the EGR valve opens to induce exhaust gas, pressure compensation between intake and exhaust port take place. In fact, this lowers pumping work, however, the largest benefit found for the highest EGR rate amounts to -0.1 g/kW in ISFC.

Figure 12 shows the dilution ratio D against the relative air-fuel ratio [lambda] and [NO.sub.x] emissions. As EGR rate increases at constant D, excess air is replaced by exhaust gas, reducing [O.sub.2] fraction and therefore [lambda]. Both [lambda] and D remain almost constant within the high efficiency regime for each EGR rate, accountable to the small influence of CA50 on engine efficiency. Across any EGR rate both [lambda] and D decrease in response to the higher fuel consumption when operating at late combustion phasing. The benefit in ISFC caused by EGR for the low [NO.sub.x] region reflects in increasing dilution ratio at same [NO.sub.x] for higher EGR rates.

Figure 13 reveals IT as function of burning delay IT-CA5 (top) and burning duration CA5-CA90 (bottom) over CA50. At late IT, the pressure difference between pre- and main combustion chamber increases, ignition conditions improve in both chambers and the engine operates at lower dilution ratio to keep IMEP constant. These three factors combined intensify the flame jets that propagate from the prechamber into the main combustion chamber, leading to reduced IT-CA5. While for the CA50 sweep without EGR burning delay increases only slightly, the gradient is significantly steeper for traces of higher EGR rates. The decelerating effect of EGR on combustion leads to longer burning duration. Starting from CA50 = 6 [degrees]CA ATDC, CA5-CA90 decreases until a combustion centre of CA50 = 10 [degrees]CA due to constant mixture dilution but improved ignition conditions. Then it rises again as the second part of the combustion occurs slower for later CA50, in spite of slightly richer mixtures.

The observed behaviour on burning delay and burning duration matches well with the results obtained from experiments with the CVCC. In either cases, combustion slows down exponentially with rising EGR rate.

Lean burn cogeneration engines conventionally operate with oxidation catalysts to convert HC and CO emissions. CO is converted readily even for low exhaust gas temperatures, whereas HC emissions represent a challenge due to higher required temperatures for conversion. The increase of burning duration for higher EGR rates correlates with exhaust gas temperature as shown as function of [NO.sub.x] emissions in Figure 14 (top). As expected, exhaust gas temperature [T.sub.exh] of every EGR rate reveals its maximum for the lowest [NO.sub.x] emissions (i.e. most retarded combustion phasing). The prolonged combustion for mixtures with EGR coupled with lower mixture dilution lead to higher exhaust gas temperatures for operating points of same CA50. This indicates a better combustion in the late part of the expansion stroke, resulting in lower HC emissions (see Figure 14, bottom). Within the high efficiency regime, the engine emits lower HC's but same [NO.sub.x] with EGR. When operating at low [NO.sub.x] emissions a different picture emerges as the engine achieves only same [NO.sub.x] with EGR for advanced combustion phasing. In conclusion, the benefit in ISFC for same [NO.sub.x] or in [NO.sub.x] for almost same ISFC goes along with slightly higher HC emissions. When comparing EGR rates of 15 and 20 % at CA50 = 16 [degrees]CA and 18 [degrees]CA with the ROP (5.7 g/kWh), HC emissions increase up to 6.2 g/kWh and 5.8 g/kWh, respectively.

Constant [NO.sub.x] Emissions

After studying the effect of EGR for constant IMEP, its influence for constant [NO.sub.x] emissions is analysed for the baseline value of 1.44 g/kWh and the half of it (0.72 g/kWh). Starting from the reference operating point of [x.sub.EGR] = 0 %, exhaust gas was recirculated and ignition timing advanced until meeting the defined value in [NO.sub.x] at an accuracy of +/- 5%. After, ignition timing was retarded and the amount of fuel increased in steps of 0.025 kg/h (lower dilution ratio D). This procedure was repeated for every EGR rate until CA50 exceeded 24 [degrees]CA ATDC.

Figure 15 shows IMEP (top) and ISFC (centre and bottom) against CA50 for all investigated EGR rates. EGR allows the engine to operate with the same amount of fuel at earlier CA50, resulting in higher IMEP as well as better ISFC for all EGR rates. The benefit, is rather small in the case of [x.sub.EGR] = 25 % due to significantly delayed combustion in combination with deteriorated cylinder filling caused by enhanced mixture temperature. This is also the reason for the significant gap in ISFC compared with the other EGR rates smaller than [x.sub.EGR] = 25 %.

For [NO.sub.x] = 1.44 g/kWh, highest advantage in engine efficiency is found at [x.sub.EGR] = 20 % as CA50 can be advanced by -4.5 [degrees]CA, improving ISFC by -4.8 g/kWh and IMEP by +0.15 bar compared with ROP. With reducing amount of fuel (i.e. lower dilution ratio D), IT needs to be retarded to obtain CA50 that still complies with the predefined [NO.sub.x] value. IMEP further increases over the whole range except for [x.sub.EGR] = 25 %. At CA50 of almost 16 CA ATDC the relative air-fuel ratio amounts to [lambda] = 1.06, which is close to the peak value with regards to [NO.sub.x] formation of [lambda] [approximately equal to] 1.1 mentioned in literature [19]. Hence, combustion phasing could be advanced beyond this value until stoichiometric combustion was reached, leading to the highest IMEP of this approach, however, at the expense of ISFC (+3.1 g/kWh).

To reduce [NO.sub.x] emissions to the half of the reference value at same CA50, the engine needs to operate without EGR at lower IMEP of 5.91 bar. This is achieved through further leaning, reducing the specific heat capacity of the exhaust gas. Along with the higher sensitivity of [NO.sub.x] emissions for very low values, the potential of EGR to increase IMEP and/or reduce ISFC diminishes for smaller [NO.sub.x] emissions. Again, [x.sub.EGR] = 20 % leads to the highest benefit in ISFC of -2.5 g/kWh (compared with [x.sub.EGR] = 0%) by advancing CA50 to almost 16 [degrees]CA ATDC. For similar ISFC the highest IMEP amounts to 6.04 bar. In correspondence with the results for higher [NO.sub.x] emissions, ISFC follows almost linear correlation to CA50 independently from EGR rate, excluding [x.sub.EGR] = 25 %.

The results show that the influence of EGR enhances with higher EGR rates and lower dilution ratio D, both increasing mixture's specific heat capacity. This becomes clear when analysing mixture's mole fraction, shown in Figure 16. For the analysis, air is divided into stoichiometric and excess part, while the recirculated exhaust gas consists only of [O.sub.2], [N.sub.2], CO and [H.sub.2]O, meaning that emissions such as [NO.sub.x], CO and HC are neglected.

In spite of sensitive actuators at the test bench, NG fraction of the analysed operating points varied slightly. Since the stoichiometric part of air is 16.28 (stoichiometric air-fuel ratio of [AFR.sub.stoich]) times the NG, even minor differences might give the impression of a large deviation emphasised by an enlarged ordinate of the diagram from 0.7 to 1.0.

For the configuration without EGR, [lambda] is below the Dilution D as residual gas is considered when calculating [lambda]. Note that Figure 16 only depicts the mixture that enters the cylinder and therefore no internal EGR fraction is shown. With augmenting EGR rate, fraction of excess air decreases and therefore [O.sub.2] concentration as well as [lambda]. The [O.sub.2] concentration of the exhaust gas increases for EGR rates up to [x.sub.EGR] = 20 %, whereas it becomes lower again for [x.sub.EGR] = 25 % due to small fraction of excess air of the yet unburnt air-NG-exhaust gas mixture. All other components of the exhaust gas increase, while merely C[O.sub.2] and [H.sub.2]O enhancing mixture's heat capacity.

Prototype Piston - CR = 15.25

After showing the potential of EGR to improve the trade-off between engine efficiency, [NO.sub.x] emissions and IMEP for the baseline compression ratio of CR = 13.20, the compression ratio was increased to CR = 15.25 by replacing the baseline with the prototype piston. The experiments were carried out in the same manner as previously described with the only difference that [x.sub.EGR] = 5 and 10 % were not considered due to their small impact on engine operating characteristics.

According to literature, combustion speeds up for higher compression ratios in response to both increased temperature level and mixture density at a given crank angle. Besides, residual gas content decreases due to the smaller in-cylinder volume at TDC during gas exchange [25]. For the presented results, 1D-CFD calculations were carried out, showing that the higher compression ratio enhances temperature by more than 30 [degrees]C (at -10 [degrees]CA ATDC) while residual gas content decreases by 0.5 %-points relative to the baseline configuration.

Constant IMEP

Figure 17 displays the trade-off between ISFC and [NO.sub.x] emissions for various EGR rates at constant IMEP = 6.35 bar. The higher compression ratio allows the engine to operate with leaner mixture to achieve IMEP and CA50 of the reference operating point. On the one hand, this improves the performance compared with the baseline configuration, but on the other, the effect of EGR decreases for higher dilution ratios as explained before (cf. Figure 9). Within the low [NO.sub.x] regime, EGR therefore improves the trade-off only moderately. The highest benefits causes an EGR rate of [x.sub.EGR] = 15 %, whereas [x.sub.EGR] = 25 % deteriorates the trade-off across the CA50 range studied. For the high efficiency regime, ISFC differs slightly with CA50 regardless of the EGR rate. [NO.sub.x] emissions increase significantly as peak combustion pressure and therefore temperature of the in-cylinder charge rise. This also favours the chance knocking combustion to takes place. The algorithm of the indicating system detected very slight knocking when operating at CA50 = 6 [degrees]CA without EGR, while this was not the case for later combustion phasing or engine operation with EGR.

Constant [NO.sub.x] Emissions

The effect of EGR on IMEP and ISFC for same [NO.sub.x] emissions reveals Figure 18. CA50 and [NO.sub.x] emissions of the reference operating point result in higher IMEP of 6.45 bar and lower ISFC of 203.0 g/kWh. When adding EGR, advanced combustion phasing becomes possible for same fraction of fuel without increasing [NO.sub.x] emissions for all investigated EGR rates. For [NO.sub.x] emissions of 1.44 g/kWh, IMEP increases slightly for EGR rates of [x.sub.EGR] = 15 % and 20 %, resulting in lower ISFC of -2.2 g/kWh. It can be seen that the achievements for the higher compression ratio are lower compared to the baseline of CR = 13.20. This results from both higher dilution ratio D and therefore smaller heat capacity of the exhaust gas as well as modified burning behaviour caused by altered piston geometry. This is even more pronounced, when operating with EGR at [NO.sub.x] emissions of 0.72 g/kWh, resulting in a worsened trade-off for almost all mixture compositions.

Influence on Knocking Combustion

The investigations at constant IMEP reveal that knocking, occurring at CA50 = 6 [degrees]CA, could be avoided by applying EGR as literature reports [13, 14, 26]. This encouraged carrying out brief trials addressing the effect of EGR on knocking combustion.

The engine was operated at maximum brake torque (MBT) found at CA50 = 8 [degrees]CA while IMEP was increased until the knock margin was met. It was reached when at least 1 of 200 cycles of multiple consecutive measurements revealed knocking combustion. Subsequently, the EGR rate was stepwise increased and ignition timing as well as dilution ratio adjusted, resulting in higher IMEP as shown in Figure 19. Knocking commences for MBT spark timing without EGR at an IMEP of 6.6 bar. The feasible value increases up to IMEP = 7.5 bar with increasing EGR rate up to [x.sub.EGR] = 20 %, (see Figure 19), while maximum pressure increases from 62 to 66 bar.

Both parameters rise exponentially due to higher EGR rate and decreasing dilution ratio (i.e. higher amount of fuel).

For the highest IMEP = 7.5 bar the engine operates at stoichiometric air-fuel ratio, exhibiting ISFC = 198.2 g/kWh. Hence, [lambda] = 1-operation including the use of a TWC might represent an alternative configuration to meet future emission standards.

Comparison of Compression Ratios at Constant IMEP

EGR improves the trade-off between engine efficiency, [NO.sub.x] emissions and IMEP for both CR = 13.20 and 15.25. While the relative potential appears to be higher for CR = 13.20, CR = 15.25 holds advantages over CR = 13.20 with respect to absolute values. To evaluate the differences between both configurations, results at constant IMEP = 6.35 bar and [x.sub.EGR] = 0 and 15 % are compared.

For engine operation without EGR, the trade-off improves with higher compression ratio over the entire investigated range (see Figure 20, top). The savings in fuel consumption caused by higher compression ratio are smaller in the high efficiency regime due to significantly higher peak combustion pressure [p.sub.max] (see Figure 20, bottom), increasing both combustion temperature and therefore wall heat losses.

The prototype piston allows the engine to operate at quite late centre of combustion of CA50 = 20 [degrees]CA ATDC, reducing [NO.sub.x] emissions by 0.3 g/kWh and even ISFC by 2.1 g/kWh (compared with the ROP). When recirculating exhaust gas of [x.sub.EGR] = 15 % and advancing CA50 to 19 [degrees]CA ATDC, [NO.sub.x] fall further to 65 % of the ROP while ISFC is still reduced by 0.6 %. The effect of EGR is relatively higher for CR = 13.20, resulting from lower dilution ratio and therefore higher specific heat capacity of the exhaust gas. This becomes clear, in particular, when comparing [NO.sub.x] emissions of operating points with same CA50 = 20 [degrees]CA (see Figure 20, top). For both compression ratios, [NO.sub.x] decrease, however, the difference is larger in the case of CR = 13.20.

The reason for a higher relative potential of EGR to improve the trade-off between efficiency and [NO.sub.x] emissions for the smaller compression ratio is not only found in higher specific heat capacity of the exhaust gas but also in differing combustion characteristics of the two pistons. Figure 21 shows burning duration CA5-CA90 as function of CA50 for both compression ratios at [x.sub.EGR] = 0 and 15 %. Starting from late combustion phasing, burning duration decreases as combustion is advanced, while it rises again as mixture becomes leaner and combustion conditions deteriorate. Combustion duration increases for both compression ratios as exhaust gas is recirculated. CA50 influences combustion duration stronger for lower compression ratios, reflected by a steeper gradient for CR = 13.20. In the case of CR = 15.25 the effect is rather marginal for the low [NO.sub.x] regime. This can be accounted for the fact that EGR allows the engine to operate at earlier CA50 regardless of the compression ratio but only for CR = 13.20 also CA5-90 decreases.

Combustion stability plays a major role for lean burn operation. High cycle-by-cycle variations lead to numerous cycles of early CA50 and thereby increasing [NO.sub.x] emissions. To analyse cycle variability, the coefficient of variation in indicated mean effective pressure [COV.sub.IMEP] is analysed for CR = 13.20/15.25 and [x.sub.EGR] = 0/15 % (see Figure 22). Lowest [COV.sub.IMEP] can be found in the high efficiency regime caused by thermodynamically more favourable CA50 and thus stabilising combustion. In turn, [COV.sub.IMEP] increases for lower [NO.sub.x] emission, when operating at late CA50. In this instance, a flatter gradient for higher compression ratio is the outcome of accelerated combustion. It also allows the engine to operate at lower cycle-by-cycle variations at [x.sub.EGR] = 15 %. For the lower compression ratio of CR = 13.20 on the other hand, EGR partly leads to small advantages at same [NO.sub.x] resulting from earlier combustion phasing. For operation at high compressions ratio, at lowest [NO.sub.x] emissions (CA50 = 20 [degrees]CA) and an EGR rate of [x.sub.EGR] = 15 % cycle-by-cycle variations are still lower than any other operating points of CR = 13.20 within the low [NO.sub.x] regime with or without EGR.

The higher compression ratio expands the mixture after combustion to a lower pressure for same CA50, resulting in a lower exhaust gas temperature. This is additionally promoted through intensified dilution in the case of higher CR = 15.25 (see Figure 23, top).

Exhaust gas temperature increases with EGR for both compression ratios due to slower combustion, in particular, in the later part of it. Within the high efficiency regime, it is consistently higher for combustion with EGR at both same [NO.sub.x] emissions. Opposing behaviour can be seen for late CA50 provoked by operating points that show same [NO.sub.x] emissions but exhibit more advanced combustion phasing. This difference is higher for CR = 13.20 due to the stronger effect of EGR. In contrary, CR = 15.25 with EGR shows lower exhaust gas temperature for same [NO.sub.x] compared to the baseline compression ratio with or without EGR, responding from earlier CA50. Since exhaust gas temperature is crucial for conversion efficiency of the oxidation catalyst, particularly for HC emissions, the choice of the employed catalyst is of vital importance when considering EGR for series engine production.

HC emissions behaves conversely to exhaust gas temperature as plotted in Figure 23 (bottom) over [NO.sub.x] emissions. EGR reduces HC emissions for both compression ratios, however, for CR = 15.25 HC emissions exceed the value of the ROP (CA50 = 19 [degrees]CA ATDC) by 0.7 g/kWh with EGR and by 1.1 g/kWh without EGR at same CA50, whereas for CR = 13.20 with EGR it drops by 0.6 g/kWh.

SUMMARY

The focus of this work was to analyse the effect of substituting a part of excess air with exhaust gas for a small lean burn NG cogeneration engine in order to achieve a better trade-off between engine efficiency, [NO.sub.x] emissions and IMEP. The outcome of this work can be summarised as follows:

* 0D calculations assuming isochoric combustion indicate potential for EGR to increase efficiency caused by a higher compression ratio without enhancing peak combustion temperature.

* Experiments with [CH.sub.4] in a constant volume combustion chamber show that mixtures of constant fuel fraction burn at lower temperatures but also slower as excess air is partly replaced with exhaust gas. This indicates the necessity to advance ignition timing for engine operation at constant combustion phasing when applying EGR.

* The C[O.sub.2] and [H.sub.2]O fraction of the exhaust gas is linked to mixture's air-fuel ratio, thus revealing a substantial effect on specific heat capacity and therefore on the efficacy of exhaust gas recirculation.

* EGR rates of 5 and 10 % showed only minor effects on engine characteristics due to small increase of specific heat capacity. As specific heat capacity increases exponentially with the EGR rate, 15 and 20 % show significant impact, leading to the highest benefits for all trials.

* For constant IMEP, the trade-off between ISFC and [NO.sub.x] emissions improves with EGR for the low [NO.sub.x] regime as a result of advanced CA50 for same [NO.sub.x] output. When operating at equal efficiency, CA50 still needs to be advanced due to prolonged combustion. [NO.sub.x] emissions (-20 %) drop the highest for EGR rates of [x.sub.EGR] = 20 % at similar ISFC when operating with CR = 13.20. A higher compression ratio improves the trade-off even without EGR caused by a faster combustion and leaner mixture. The additional benefit is relatively small when adding exhaust gas to the mixture. In fact, for engine operation at [x.sub.EGR] = 15 %, ISFC and [NO.sub.x] emissions decrease by 0.2 % and by 35 %, respectively, compared with the reference operating point. The high efficiency regime suffers by applying EGR, mainly attributable to the small dependency of ISFC on CA50 for early combustion phasing. The loss caused by delayed combustion outweighs the benefit achieved for advanced CA50.

* For constant [NO.sub.x] emissions of the reference operating point, EGR allows the engine to operate at advanced CA50 of -4.5 [degrees]CA, leading to a plus in IMEP of 0.15 bar coupled with lower fuel consumption of -4.7 g/kWh. A higher compression ratio reduces ISFC benefit by another -0.9 g/kWh at same IMEP. EGR shows rather small potential for halved [NO.sub.x] emissions due to lower specific heat capacity and modified piston geometry.

* Very slight knocking occurs for CR = 15.25 at early CA50, which can be avoided through EGR. Experiments at MBT show that EGR enhances the knock margin from IMEP = 6.6 without EGR to 7.5 bar by adding [x.sub.EGR] = 20 %. The engine then operates with stoichiometric mixture, which is particularly valuable with respect to possible necessity for [lambda] = 1.0-concepts with TWC.

The findings show that EGR represents a considerable submeasure for future cogeneration engines concepts to comply with progressively stringent legislated emission limits. In spite the fact that no premature wear could be seen during and after engine trials, investigations with this regard over a longer period appear to be necessary to ensure reliable engine operation in the field. The same applies to the effect of fluctuating fuel composition as well as varying ambient conditions. Another challenge that needs to be faced is condensation of the exhaust gas during warm up time of the system, which could be avoided by increasing exhaust gas temperature for the very first part of engine operation.

REFERENCES

[1.] European Union, "Energy Roadmap 2050," Brussels, 2011.

[2.] Kerr, T., "Combined Heat and Power: Evaluating the Benefits of Greater Global Investement," Paris, 2008.

[3.] Bundesministerium fur Umwelt, Naturschutz und Reaktorsicherheit, "Technische Anleitung zur Reinhaltung der Luft: TA Luft," 2002.

[4.] European Union, "Commission Regulation (EU) No 813/2013 of 2 August 2013 implementing Directive 2009/125/EC of the European Parliament and of the Council with regard to ecodesign requirements for space heaters and combination heaters," 2013.

[5.] Wimmer, A., "Gas Engine versus Diesel Engine: A Comparison of Efficiency," MTZ industrial 1(1):22-27, 2011.

[6.] Wimmer, A., Pirker, G., Zelenka, J., and Chmela, F., "The Potential of Exhaust Gas Recirculation in Large Gas Engines," Shanghai, China, 2013.

[7.] Nellen, C. and Boulouchos, K., "Aufgeladene Gasmotoren mit AGR und Dreiwege-Katalysator - der Weg zu niedrigsten Emissionen bei hohem Wirkungsgrad und gro[beta]er Leistungsdichte," MTZ 61(1):54-62, 2000.

[8.] Doosje, E., "Limits of mixture dilution in gas engines," Ph.D. Thesis, Eindhoven University, 2010.

[9.] Einewall, P., Tunestal, P., and Johansson, B., "Lean Burn Natural Gas Operation vs. Stoichiometric Operation with EGR and a Three Way Catalyst," SAE Technical Paper 2005-01-0250, 2005, doi:10.4271/2005-01-0250.

[10.] Yan, B., Yao, M., Mao, B., Li, Y. et al., "A Comparative Study on the Fuel Economy Improvement of a Natural Gas SI Engine at the Lean Burn and the Stoichiometric Operation both with EGR under the Premise of Meeting EU6 Emission Legislation," SAE Technical Paper 2015-01-1958, 2015, doi:10.4271/2015-01-1958.

[11.] Lee, S., Ozaki, K., Iida, N., and Sako, T., "A Potentiality of Dedicated EGR in SI Engines Fueled by Natural Gas for Improving Thermal Efficiency and Reducing NOx Emission," SAE Int. J. Engines 8(1):238-249, 2015, doi:10.4271/2014-32-0108.

[12.] Pizzirani, N., Lauer, T., Geringer, B. Murakami, S., "A Numerical Method for Efficiency Optimization on Gas Engines," MTZ industrial 1(2):60-67, 2012.

[13.] Tschalamoff, T., "Abgasruckfuhrung im mittelschnell laufenden Gasmotor," MTZ 65(11):932-939, 2004.

[14.] Tschalamoff, T. and Kauert, L., "Potenziale der Abgasruckfuhrung im mittelschnelllaufenden Gasmotor mit offenem Brennraum und Vorkammer," 3rd Dessau Gas Engine Conference, Dessau, Germany, 2003.

[15.] Bhargava, S., Clark, N., and Hildebrand, M., "Exhaust Gas Recirculation in a Lean-Burn Natural Gas Engine," SAE Technical Paper 981395, 1998, doi:10.4271/981395.

[16.] Peer, J., Wachtmeister, G., and Boulouchos, K., "Einfluss der Abgasruckfuhrung bei Magerkonzept-Gasmotoren mit extremen Steuerzeiten auf Brennverlauf und Schadstoffemissionen unter Berucksichtigung der Gasqualitat," FVV Informationstagung, Leipzig, Germany, 2013.

[17.] Schoffler, T., Hoffmann, K., and Koch, T., "Stoichiometric Natural Gas Combustion in a Single Cylinder SI Engine and Impact of Charge Dilution by Means of EGR," SAE Technical Paper 2013-24-0113, 2013, doi:10.4271/2013-24-0113.

[18.] Tinaut, Fluixa, F. V., Gimenez, Olavarria, B., Iglesias, Hoyos, D. Lawes, M., "Experimental Determination of the Burning Velocity of Mixtures of n-Heptane and Toluene in Engine-like Conditions," Flow, Turbulence and Combustion(89):183-213, 2012.

[19.] Heywood, J., "Internal Combustion Engine Fundamentals," vol. 1988, McGraw-Hill, ISBN 0-07-028637-X, 1988.

[20.] Tinaut, Fluixa, Francisco V. , Melgar, A., Gimenez, Olavarria, B. Reyes, M., "Characterization of the combustion of biomass producer gas in a constant volume combustion bomb," Fuel(89):724-731, 2010.

[21.] Lafuente, A., "Metodologia para el diagnostico de la velocidad de combustion laminar de mezclas de gases combustibles a partir de la medida de la presion instantanea en una bomba de combustion a volumen constante," Ph.D. Thesis, Universidad de Valladolid, 2008.

[22.] Iglesias, D., "Metodologia para el estudio de propiedades de combustion de sustancias liquidas a partir de ensayos en bomba de combustion a volumen constante," Ph.D. Thesis, Universidad de Valladolid, 2008.

[23.] Keppeler, R., "Entwicklung und Evaluierung von Verbrennungsmodellen fur die Large Eddy Simulation der Hochdruck-Vormischverbrennung," Ph.D. Thesis, Universitat der Bundeswehr Munchen, 2013.

[24.] Gu, X.J., Haq, M.Z., Lawes, M. Woolley, R., "Laminar Burning Velocity and Markstein Lengths of Methane-Air Mixtures," Combustion and Flame 121(1-2):41-58, 2000.

[25.] Das, A. and Watson, H.C., "Development of a Natural Gas Spark Ignition Engine for Optimum Performance," Proc. of Inst. Mech. Eng. Part D: J. of Automob. Eng., 211(5):361-378, 1997.

[26.] Nellen, C. and Boulouchos, K., "Natural Gas Engines for Cogeneration: Highest Efficiency and Near-Zero-Emissions through Turbocharging, EGR and 3-Way Catalytic Converter," SAE Technical Paper 2000-01-2825, 2000, doi:10.4271/2000-01-2825.

CONTACT INFORMATION

Denis Neher

Research Associate

Karlsruhe University of Applied Sciences

Moltkestra[beta]e 30

76133 Karlsruhe

Germany

denis.neher@hs-karlsruhe.de

ACKNOWLEDGMENTS

The German Federal Ministry of Education and Research supported this project through the grant programme "Ingenieurnachwuchs". The authors would like to thank all persons involved. They also like to thank the AVL List GmbH for providing the simulation software BOOST.

DEFINITIONS/ABBREVIATIONS

A/V - Surfac area to volume ratio

ABB - Arbeitsstelle Bertolt Brecht

Af - Flame front

[AFR.sub.stoich] - Stoichiometric air-fuel ratio

[A.sub.pb] - Surface area of piston bowl

ATDC - After top dead centre

AVL - Anstalt fur Verbrennungs-kraftmaschinen List

BDC - Bottom dead centre

BMEP - Brake mean effective pressure

CA - Crank angle

Cc - Laminar burning velocity

C[H.sub.4] - Methane

CO - Carbon monoxide

C[O.sub.2] - Carbon dioxide

CR - Compression ratio

[C.sub.sq] - Squish ratio

CU - Cogeneration unit

[c.sub.v] - Specific heat capacity at constant volume

CVCC - Constant volume combustion chamber

D - Dilution ratio

[dm.sub.b]/dt - Mass burning rate within combustion bomb

EGR - Exhaust gas recirculation

[H.sub.2]O - Water

HC - Hydrocarbon

HV - Higher heating value

IMEP - Indicated mean effective pressure

ISFC - Indicated specific fuel consumption

IT - Ignition timing

MBT - Maximum brake torque

[m.sub.fuel] - Mass of fuel added to the cycle

n - Engine speed

[N.sub.2] - Nitrogen

[n.sub.air] - Number of moles of air

[n.sub.air stoich] - Number of moles of air required for stoichiometric combustion

[n.sub.EGR] - Number of moles of recirculated exhaust gas

NG - Natural gas

[n.sub.O2 stoich] - Number of moles of [O.sub.2] required for stoichiometric combustion

[O.sub.2] - Oxygen

[p.sub.0] - Initial pressure

[p.sub.CU in] - Inlet pressure of cogeneration unit

[P.sub.el] - Electrical power

[p.sub.max] - Peak combustion pressure

[p.sub.max] - Peak in-cylinder pressure

ROP - Reference operating point

[T.sub.0] - Initial temperature

Tb - Temperature of burnt zone

[T.sub.CU in] - Inlet temperature of cogeneration unit

TDC - Top dead centre

[T.sub.exh] - Exhaust gas temperature

Tu - Temperature of unburnt zone

TWC - Three way catalyst

[V.sub.BDC] - Volume at bottom dead centre

[V.sub.TDC] - Volume at top dead centre

[x.sub.EGR] - EGR rate

[lambda] - Relative air-fuel ratio

[[rho].sub.u] - Density of unburnt mixture

([phi]) - Fuel-air equivalence ratio

Denis Neher, Fino Scholl, and Maurice Kettner Karlsruhe University of Applied Sciences

Danny Schwarz and Markus Klaissle Senertec Kraft-Warme-Energiesysteme GmbH

Blanca Gimenez Olavarria University of Valladolid

Table 2. Typical natural gas compositions during experiments

                       Concentration [vol %]
Substance                  Day 1   Day 2

Methane                   92.720  93.451
Ethane                     4.640   4.410
Nitrogen                   0.975   0.857
Carbon dioxide             0.878   0.754
Propane                    0.584   0.388
Butane (iso, n)            0.166   0.120
Pentane (neo, iso, n)      0.034   0.016
n-Hexane                   0.003   0.003
COPYRIGHT 2016 SAE International
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2016 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Neher, Denis; Scholl, Fino; Kettner, Maurice; Schwarz, Danny; Klaissle, Markus; Olavarria, Blanca Gi
Publication:SAE International Journal of Engines
Article Type:Report
Date:Dec 1, 2016
Words:10953
Previous Article:Alternative engine speed sensing using the electric signals of the alternator.
Next Article:Development of hydraulic-controlled variable valve lift system for scooter engine.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters