Methane Emissions from Natural Gas Vehicles in China

The study investigates whether the rapid adoption of natural gas vehicles (NGVs) in China delivers greenhouse gas (GHG) benefits once methane (CH4) emissions from vehicles are accounted for. Although NGVs have been promoted for air quality benefits (lower PM2.5 and NOx) and lower CO2 per unit energy, vehicle CH4 emissions in China have not been well quantified and have often been omitted from inventories and life-cycle assessments. The research questions are: (1) What are real-world CH4 emission factors (EFs) for Chinese NGVs, especially heavy-duty buses and trucks? (2) How do these CH4 emissions affect well-to-wheels (WTW) GHG impacts compared with gasoline and diesel vehicles? (3) How might enforcement of the China VI standard alter future CH4 and GHG emissions? The study is important because China has the world’s largest NGV fleet, heavy-duty NGVs are growing rapidly, and real-world CH4 emissions could negate climate benefits if high.

Prior Chinese and international life-cycle analyses often emphasized upstream natural gas (NG) extraction, processing, and distribution emissions, while largely neglecting vehicle CH4 emissions due to lack of measurements in China. IPCC default vehicle CH4 EF is 0.4% of NG consumed, but Hu et al. (2018) reported ~3.0 ± 0.5% overall on-road CH4 EF for Chinese NGVs, suggesting much higher emissions. Many Chinese NGVs are retrofits (about 80%), potentially increasing CH4 due to engines and after-treatment not optimized for NG. Heavy-duty Chinese NGVs commonly use lean-burn (LB) engines with oxidation catalysts (OC), which are less effective for CH4 oxidation at typical exhaust temperatures than stoichiometric (SM) engines with three-way catalysts (TWC). US studies (e.g., Clark et al.) indicated vehicle operation can dominate pump-to-wheels CH4 emissions, and well-to-wheels CH4 can be 40–60% from vehicle operation. However, Chinese vehicle CH4 EFs—especially for heavy-duty vehicles—were missing, and compliance tests (ETC) may not reflect real-world lower exhaust temperatures that diminish CH4 removal efficiency. This study addresses these gaps with direct on-road measurements and integrated life-cycle and inventory analyses.

Field measurements: During the 2014 CAREBEIJING North China Plain campaign, a mobile laboratory with fast-response sensors (10 Hz) performed on-road plume-chasing in Baoding and Shijiazhuang to quantify CH4 emissions from NG buses and taxis. Approximately 26 hours of driving (~600 km) captured emissions from 73 NG buses and 63 NG taxis. Plumes were identified using enhanced CH4 and CO2 signals, temporal co-variation, and video confirmation. Interference from nearby vehicles was minimized with correlation filtering (R^2 ≥ 0.5 within ±1 s) and orthogonal regression of CH4 and CO2 enhancements. A Gaussian puff model was used to evaluate method robustness against multi-source plumes. Emission ratio and EF derivation: The CH4:CO2 emission ratio (ER) was computed from orthogonal regression slopes of enhancements (ΔCH4/ΔCO2). Fuel-specific CH4 EF (% of NG consumed) was derived via carbon balance: EF_CH4,obs = ER/(1+ER) × 100%. Sensitivities to thresholds (ΔCH4, ΔCO2, R^2) were assessed; major uncertainty stemmed from plume identification (quantified via variations with/without a CH4 cutoff of 0.4 ppm). Adjustments: Observed EFs were adjusted for (1) cold-start emissions using a cold-start/hot-start ratio of 1.5 with 14% cold-start weighting (per China VI test procedure), and (2) sporadic tank venting emissions not captured by co-emission methods by adding 0.1% of NG consumed, based on literature. Upper-bound uncertainties considered stronger cold temperature effects. Heavy-duty NG trucks: Direct identification in the field was difficult; thus, bus EFs were used as a proxy for heavy-duty NG trucks equipped with similar LB engines + OC in China. Additional uncertainty was assigned to reflect potential differences in highway operation. Life-cycle and WTW analysis: WTW GHG emissions combined upstream (well-to-pump, WTP) GHG EFs for NG, vehicle fuel consumption, and adjusted CH4 EFs. WTP CH4 leakage rates of 1.65 ± 1.05% of NG consumed and WTP GHG EF of 28 ± 6 g CO2eq MJ−1 were used for both CNG and LNG, as differences due to transport distances were smaller than distribution leakage variability. Bottom-up CH4 inventory (2000–2017): NG consumption for four categories (NG taxis, private light-duty NGVs, NG buses, heavy-duty NG trucks) was estimated from vehicle population, annual mileage traveled, and distance-specific fuel consumption. Category-specific adjusted CH4 EFs were applied to estimate total CH4 emissions. Consistency with national statistics (China Statistical Yearbook Transport, Storage, and Post sector) was checked, noting exclusions (e.g., private vehicles not included in CSYB sector figures, cargo ships). Scenario analyses (2020–2030): Three scenarios evaluated impacts of China VI implementation and fleet evolution: high-emission (continued retrofits; LB+OC dominance; loose CH4 enforcement), medium-emission (no retrofits; SM or HPDI adoption; moderate penetration), and low-emission (real-world CH4 at China VI limits; NGV growth localized to low-leakage regions; lower distribution leakage). Vehicle population projections, fuel consumption improvements (15% after 2021 per Stage 3 limits), and technology pathways were incorporated.

- Measured on-road fuel-specific CH4 EFs: light-duty NG taxis averaged 1.7 ± 0.5% of NG consumed; heavy-duty NG buses averaged 2.9 ± 0.5%. - LNG vs CNG buses showed no significant EF difference (LNG: 2.8 ± 0.4%; CNG: 3.1 ± 0.5%). Buses used LB engines with OC; low NH3 emissions corroborated LB+OC behavior. - The observed CH4 EF for NG buses was about 90% higher than the China V CH4 limit for heavy-duty vehicles, indicating substantial real-world exceedances. - Seasonality and venting-adjusted EFs: taxis 1.9% [−0.7, +0.9], buses 3.2% [−0.8, +1.0], and trucks assumed 3.2% [−1.7, +1.0] based on bus EFs and uncertainties. - WTP NG leakage rate used: 1.65 ± 1.05% of NG consumed; WTP GHG EF: 28 ± 6 g CO2eq MJ−1. - WTW impacts: With observed CH4 EFs, NG cars and buses generally do not reduce GHG compared to gasoline cars and diesel buses; current-generation NG trucks with LB+OC likely increase GHG by about 160 [−200, +180] g CO2eq km−1 versus diesel. If China VI CH4 limits are achieved in real-world operation, NG trucks could reduce GHG by ~100 ± 150 g CO2eq km−1, and upstream leakage would become the limiting factor. - Historical totals: Annual CH4 emissions from NGVs rose from ~0.0014 [−0.0004, +0.0004] Mt in 2000 to ~0.77 [−0.28, +0.22] Mt in 2017; >80% in 2017 came from buses and trucks. Switching to NGVs increased cumulative WTW GHG emissions by about 83 Mt CO2eq over 2000–2017. - Future scenarios (2020–2030): • High-emission: CH4 reaches ~3.3 Mt y−1 by 2030; cumulative WTW GHG change +432 Mt CO2eq. • Medium-emission: CH4 ~1.3 Mt y−1 by 2030; cumulative +117 Mt CO2eq. • Low-emission: CH4 decreases to ~0.7 Mt y−1 by 2030; cumulative −77 Mt CO2eq. • Comparing high vs low: stringent China VI enforcement could deliver ~509 Mt CO2eq GHG reduction (2020–2030) and ~70% lower CH4 in 2030 than the high-emission pathway.

Real-world CH4 emissions from Chinese NGVs—especially heavy-duty buses (and by proxy, trucks with similar LB+OC technology)—are substantially higher than regulatory limits and prior inventory assumptions. Consequently, the rapid adoption of NGVs has not yielded climate benefits and has likely increased overall GHG emissions to date. The analysis shows that heavy-duty segment performance dominates fleet-level CH4 emissions and WTW climate impacts because of higher fuel use and higher CH4 EFs. Achieving real-world CH4 emissions at or below China VI limits is pivotal: under such conditions, NG trucks can provide net GHG reductions relative to diesel, shifting the key constraint to upstream leakage. The work highlights shortcomings of prior test cycles (ETC) that over-represent high exhaust temperatures and understate real-world CH4; the China VI move to WHSC and PEMS should better reflect real-world operation, though crankcase and venting emissions may still be under-captured. Policy implications include the need for stringent certification and in-use compliance, closed crankcase ventilation for SM+TWC heavy-duty NGVs, and technology development (e.g., improved CH4 oxidation catalysts) if LB operation persists. The methods demonstrated (plume-chasing and fleet-level inference) can support compliance verification and inventory improvements and are transferable to other regions with large NGV fleets (e.g., India).

This study provides the first extensive real-world CH4 emission measurements for Chinese NGVs and integrates them into life-cycle and national emission inventory assessments. Key contributions include quantifying high on-road CH4 EFs for heavy-duty NG buses (and by extension trucks), demonstrating that NGV adoption in China increased net GHG emissions since 2000, and showing via scenarios that stringent China VI implementation could reverse this trend and deliver substantial GHG benefits (up to ~509 Mt CO2eq reduction in 2020–2030). Future research should focus on: (1) expanded, seasonally resolved measurements across regions and vehicle classes, including direct measurements of heavy-duty NG trucks; (2) improved characterization and mitigation of crankcase and venting emissions; (3) verification of real-world compliance under China VI (including PEMS enhancements); and (4) development and deployment of more effective CH4 oxidation catalysts and engine control strategies that ensure low CH4 emissions without sacrificing fuel economy or NOx control.

- Limited direct measurements for heavy-duty NG trucks; truck EFs inferred from buses with similar LB+OC technology, with added uncertainty for different duty cycles (e.g., highway operation). - Possible underestimation of CH4 due to lack of cold-season data and potential omission of sporadic tank venting events (partly addressed via adjustments and literature-based venting term). - National inventory uncertainties due to incomplete vehicle population stratification, NG consumption data, and regional variability in NG distribution leakage and vehicle usage. - Test cycle representativeness: historical ETC overestimates exhaust temperatures relative to urban operation; even with China VI WHSC and PEMS, crankcase and venting emissions may not be fully captured, and PEMS results can be variable. - Scenario outcomes depend on assumptions about technology adoption, enforcement stringency, and upstream leakage rates; real-world implementation may differ.