Emission Reductions of Carbonyl Compounds in a Heavy-Duty Diesel Engine Supplemented with H 2 / O 2 Fuel

Carbonyl compounds play an important role in atmospheric chemistry, and have adverse effects on human health because they are precursors of ozone and peroxyacyl nitrates. This study investigated the emissions of carbonyl compounds from a heavy-duty diesel engine (HDDE) at one low load steady-state condition (24.5% of the max load at 40 km/h). The experimental results indicate that the emissions of total carbonyl concentrations decreased by 4.7% to 32.3% when 10 to 70 L/min H2/O2 mixture was added to neat diesel fuel. The emissions of nine individual carbonyl species also decreased. Among these, formaldehyde was the major species, accounting for 80.4%–81.5% of the total concentrations in the exhausts of all tested fuels, and its emissions were reduced from 5.1% to 31.7%. Meanwhile, the emission factors of total carbonyl compounds, in mg/L, decreased from 3.6% to 13.0%, and those of total carbonyl compounds, in mg/kWh, decreased from 4.6% to 32.3%. These results indicate that the addition of a H2/O2 mixture to neat diesel fuel can reduce emissions form diesel engines to the atmosphere.

Moreover, formaldehyde and acetaldehyde are toxic contaminates, mutagens, and carcinogens (Goldmacher and Thilly, 1983;Shepson et al., 1986;IARC, 2004).To reduce their emission levels from diesel engines are therefore desirable for human health and environment.
One way to reduce emissions of air pollutants from diesel engines is to use clean fuel.Hydrogen is a widely acknowledged as a renewable, recyclable and clean fuel.Compared to hydrocarbon fuels, hydrogen fuel has wider flammability limits, higher flame speed and faster burning velocity (Verhelst et al., 2005;Bari and Mohammad, 2010), which enable engines running on very lean mixtures (Verhelst and Sierens, 2001;Bari and Mohammad, 2010).Wang et al. (2011) found that the engine performance and energy saving were improved when hydrogen/oxygen fuel was injected into the combustion chamber of an HDDE, accompanying with emission reductions of total hydrocarbons, carbon monoxide, carbon dioxide, and nitrogen oxides; emission reductions of particulate matter and PAHs were also reported in Wang et al. (2012).
This study continued our previous work to replace the sampling system to allow collecting carbonyl compounds in an HDDE supplemented with various amounts of H 2 /O 2 fuel.The main goals here were to analyze the emission concentrations and emission reductions of total and nine individual carbonyl species.Emission factors of total and nine carbonyl species were also examined.

Test Engine and Hydrogen/Oxygen Fuel
Table 1 lists the engine specifications used in this work, which was the same as in our previous works (Wang et al., 2011;Wang et al., 2012).The direct-injection, non-catalyst, heavy-duty diesel engine was a Cummins B5.9-160, containing six cylinders with fuel injection sequence 1-5-3-6-2-4.The engine had a bore 102 mm in diameter and a stroke of 120 mm.The total displacement was 5883 mL and the compression ratio was 17.9:1.The maximum horsepower was 118 kW at 2500 rpm, and the maximum torque was 534 Nm at 1600 rpm.
A report from Taiwan's Ministry of Transportation and Communications indicated that the mean speed of vehicles in urban areas ranges from 28.7 to 45.4 km/h, with an average of 38.0 km/h (MOTC, 2008).Therefore, the engine was tested in a dynamometer (Schenck GS-350 Model) at one low load steady-state condition, corresponding to 24.5 to 28.0% of the maximum load (about 40 km/h).Fig. 1 shows the schematic of the experimental setup.
In this study, an oxy-hydrogen generator machine (Epoch EP-560A Model) was used to electrolyze water to generate hydrogen and oxygen (H 2 /O 2 ) mixture.The H 2 /O 2 was then directed to the combustion chamber of the test engine as a supplemented fuel.A gas flow meter was used to measure the flow rate of H 2 /O 2 .Two flame arrestors were installed in the H 2 /O 2 line to suppress explosions before the mixture was transported to the engine via the air inlet manifold.The constituents of H 2 /O 2 were not further analyzed here.
Although adding H 2 /O 2 can save fuel due to improved combustion efficiency, but it requires additional electricity to electrolyze water to produce hydrogen/oxygen gases.In  (Wang et al., 2012a), the electricity needed to electrolyze water to produce H 2 /O 2 was converted to diesel-fuel equivalent, and total fuel consumption, i.e., the brake specific fuel consumption (BSFC), is the sum of diesel fuel and diesel-fuel equivalent.Wang et al. (2012a) found that for 10 to 40 L/min of H 2 /O 2 mixture addition, the BSFC was higher than that of neat diesel.However, for 50, 60 and 70 L/min of H 2 /O 2 mixture addition, the BSFC was lower than that of neat diesel by about 3.2%, 9.9% and 10.5%, respectively; hence, both fuel and energy saving can be achieved.

Sampling
The engine exhausts were sampled through a critical flow Venturi-type (CFV) wind tunnel, 350 mm in diameter (Wang et al., 2011;Wang et al., 2012).After passing through the tunnel, the exhausts were diluted with air simultaneously drawn into the tunnel, and then mixed completely by a Spencer blower.All exhausts were introduced into the dilution system through a solid insulated pipe, 10 cm in diameter and 7.5 m in length.The exhaust streams were collected in a 10 L Tedlar bag (SKC-10L) placed inside a vacuum sampling box (SKC-40L) according to US EPA Method 18.Air was pumped out of the vacuum box, thus causing the exhausts to be sucked into the Tedlar bag.The total sampling time for each run lasted more than 10 min.After sampling, all Tedlar bags were stored in black opaque plastic bags to keep them away from sunlight.The samples collected in Tedlar bags were immediately pumped through a 2,4-DNPH-coated cartridge (Sep-Pak cartridge, Supelco) at a flow rate of 200 mL/min for 10 min.All carbonyl compounds captured in the cartridge were subsequently converted to corresponding hydrazone derivatives.Cartridges were then eluted with acetonitrile.The eluted solutions were analyzed using a high performance liquid chromatograph (Hewlett Packard 1100 series HPLC) with an ultraviolet-visible detector at λ = 360 nm.Filters with active carbons were used to clean the ambient dilution air.Clean ambient air was used to dilute and lower the temperature of the original exhausts.The sampling time was 20 min per test run.

Analysis
The sampling method described above allowed analyses of carbonyl compounds within 1-2 h.Nine carbonyl compounds, including formaldehyde, acetaldehyde, acrolein, acetone, propionaldehyde, crotonaldehyde, butyraldehyde, valeraldehyde, and benzaldehyde, were identified and quantified with an Agilent HP 1100 HLPC/UV.These nine carbonyl compounds were chosen since they were the dominant carbonyl species associated with diesel engine emissions (Kean et al., 2001;Kristensson et al., 2004;Legreid et al., 2007).The standard solutions of the nine carbonyl compounds were purchased from Rescek.A column C 18 (5 μm in film thickness, 4.6 mm i.d., and 250 mm in length; Zarbox, USA) was used to separate the nine compounds.Solvent A (water/acetonitrile/tetrahydrofuran mixture at 60:30:10 by volume) and/or solvent B (water/ acetonitrile mixture at 40:60 by volume) were used as the mobile phase.The flow rate was maintained at 1.5 mL/min and the injection volume was 20 mL.The concentration gradients of solvents A and B in the mobile phase were varied linearly from 100% solvent A at the beginning to 100% solvent B after 10 min.\For the blank tests, ultrapure nitrogen was introduced into a Tedlar bag.The content of the bag was then analyzed for carbonyl compounds in the same procedures described earlier.The amounts of formaldehyde and acetaldehyde measured from the blank test accounted for 0.50% and 0.92% of a typical sample, respectively.Field blank tests show that the amounts of formaldehyde and acetaldehyde per cartridge were 0.035 and 0.05 μg, respectively.
To determine the detection limit (DL) of carbonyl compounds, seven measurements were made on a sample whose concentration was at or near the detection limit.The standard deviation and variance were calculated for the data set and used to calculate the DL at the 99% confidence level.The DLs of the nine carbonyl compounds were determined as between 0.112 and 0.283 μg.Some researchers (Tejada, 1986;Possanzini and Dipalo, 1995;Risner, 1995;Mohammadi et al., 2007) reported that several dimmers form when acrolein and crotonaldehyde are absorbed onto 2,4-DNPH cartridges.In this study, the recovery efficiencies for carbonyl compounds were determined using the same procedures for the blank tests as described earlier, except that the bag was spiked with a know as n amount of each individual aldehyde.The results show that the recovery efficiency of carbonyl compounds varied from 97.4% to 106.1%, with an average of 102.6%.The mean relative standard deviation in recovery efficiencies was below 14.2%.The breakthrough test was conducted by pumping the exhaust collected in the Tedlar bags out and through two 2,4-DNPH cartridges connected in series.The measured carbonyl compounds in the front cartridge were greater than 95%, while those for the back cartridge were less than 5%.To cope with the wide concentration ranges of carbonyl compounds emitted from the diesel engine exhaust, calibration curves for both high and low concentrations of carbonyl compounds were prepared during each sampling day.

Carbonyl Concentrations in the Exhausts of HDDE
Table 2 lists the measured concentrations of nine carbonyls from the engine exhausts.Total concentration of carbonyl compounds using neat diesel was 3218 μg/m 3 .These were 3068, 3006, 2823, 2707, 2501, 2217, and 2178 μg/m 3 for neat diesel fuel supplemented with 10 to 70 L/min, at intervals of 10 L/min of H 2 /O 2 mixture, corresponding to emission reductions of 4.7%, 6.6%, 12.3%, 15.9%, 22.3%, 31.1%, and 32.3%, respectively, when 10 to 70 L/min H 2 /O 2 mixtures were added to the neat diesel.That is, total carbonyl emissions from the diesel engine decreased with increasing H 2 /O 2 mixtures.

CONCLUSIONS
The experimental results indicate that the emissions of total carbonyl concentrations were decreased by 4.7% to 32.3% when supplemented 10 to 70 L/min H 2 /O 2 mixture with neat diesel fuel.The emissions of nine individual carbonyl species were also decreased.Among these, formaldehyde was the major species, accounting for 80.4%-81.5% of total concentrations in the exhausts for all tested fuels, and its emissions were reduced from 5.1% to 31.7%.Meanwhile, the emission factors of total carbonyl compounds, EF total-carbonyls in mg/L, were decreased from 3.6% to 13.0%, and EF total-carbonyls in mg/kWh were decreased from 4.6% to 32.3%.These results indicate that neat diesel fuel supplemented with H 2 /O 2 mixture can reduce air pollutant emissions form diesel engines to the atmosphere.

Fig. 1 .
Fig. 1.The schematic of experimental setup, including wind tunnel, diesel engine, dynamometer, and sampling system.

Fig. 2 .
Fig. 2. Carbonyl compound profiles in the exhausts of the HDDE mixed with various H 2 /O 2 flow rates.

Table 1 .
The diesel engine specifications.

Table 2 .
Carbonyl compound concentrations in the exhausts of the HDDE mixed with various H 2 /O 2 flow rates (unit: μg/m 3 ).

Table 3 .
Carbonyl compound emission factors in mg/L in the exhaust of the HDDE fueled with various H 2 /O 2 flow rates.

Table 4 .
Carbonyl compound emission factors in mg/kWh in the exhaust of the HDDE fueled with various H 2 /O 2 flow rates.