Optimal Speed Ranges for Different Vehicle Types for Exhaust Emission Control

The study addresses how vehicle speed influences exhaust and carbon emissions on expressways and seeks to determine vehicle-type-specific optimal speed ranges that minimize emissions while maintaining efficient transport. Although emission standards and after-treatment technologies have advanced (Euro, EPA/LEV, China, ADR; TWC, GPF, DOC, SCR, DPF, ASC), speed limits are typically set for safety, not environmental objectives. Prior work shows a U-shaped relation between speed and emissions and that stable, non-aggressive driving reduces emissions, but most research focuses on urban/suburban settings, not highways. With the advent of variable speed limits, integrating environmental goals into speed management is a key gap. This study aims to quantify emission characteristics (CO, NOx, PM2.5, HCs) for light passenger (M1) and freight (N1–N4) vehicles across speeds, using controlled experiments and a calibrated MOVES model, to propose optimal highway speed ranges for emission control and sustainability.

The paper reviews international emission standards (EU Euro with RDE focus; US EPA and California LEV prioritizing NMOG and PM; China I–VI aligning with EU; Australia ADR akin to Euro 5) and technologies (TWC, GPF, DOC, SCR, DPF, ASC). It highlights recent engine advances (e.g., extreme Miller cycle in heavy-duty diesel engines) and control technologies reducing pollutants. The speed–emission relationship is complex: both very low and very high speeds can elevate emissions due to suboptimal combustion or aerodynamic drag/engine inefficiency, supporting a U-shaped pattern. Driving behavior studies show aggressive maneuvers raise emissions, while eco-driving reduces them, but findings largely stem from urban/suburban contexts. Variable speed limits are increasingly used for safety and flow; few studies set speed limits explicitly to minimize emissions. Research gaps include vehicle-type-specific highway speed recommendations targeting pollutant and carbon reductions, and integrating single-vehicle behavior with traffic flow effects for policy and intelligent traffic systems.

Approach: The study targets representative pollutants CO, NOx, PM2.5, and HCs for gasoline M1 light passenger vehicles and diesel N freight vehicles. It combines controlled single-vehicle exhaust tests with traffic flow emission simulations using MOVES, calibrated with experimental data, to generalize findings to highway traffic. Single-vehicle exhaust emission test: Conducted indoors on a chassis dynamometer to isolate speed effects, maintaining stable speeds while measuring emissions with a KANE AUTO5-1 five-gas analyzer (CO, HCs, O2, CO2, NO; oil temperature; engine speed; storage up to 250 results). Vehicle selection reflects Chinese expressway composition dominated by M1 light passenger vehicles. Test vehicles and key specs: (1) M1 passenger—Volvo XC40, GVW 1757 kg, gasoline, China 6 (GB18352.6-2016), 2.0 L turbo, 140 kW, TWC + GPF, automatic/electric transmission (Aisin). (2) N2 freight—Dongfeng DFL1160B, GVW 12,000 kg, diesel, China 5 (GB17691-2005), 6.5 L turbo, 132 kW, SCR + DPF, manual transmission. Speed ranges reflect highway operating data: M1 tested at 60–130 km/h; N trucks at 40–100 km/h. Traffic composition data from several expressways and operating speed statistics (by lane counts) informed speed ranges. Traffic flow emissions via MOVES: MOVES2014 Project (micro-level) mode used; VSP-based emissions; localized and calibrated with Chinese conditions and single-vehicle results. Key settings: year 2009 (aligned to China-6 vs US standard mapping), road type rural restricted access (expressway), geographic proxy Fulton County, Georgia (similar climate), section length 65 km, traffic volume 3500 veh/h, vehicle/fuel categories: light/medium passenger gasoline, large passenger diesel, minivan gasoline, freight diesel, freight train diesel; composition: light passenger 0.85, medium passenger 0.01, large passenger 0.03, minivan 0.02, freight vehicle 0.06, freight train 0.03. Inputs included meteorology, fleet age, fuels, road and operating conditions, and speed. MOVES non-exhaust PM module was also enabled to estimate brake/tire/road dust PM for PM2.5/PM10. Calibration used experimental emission characteristics to adjust parameters. Emission outputs were generated for speeds typically from 50 to 130 km/h in 10 km/h steps for each vehicle class (M1 and N1–N4).

- Emission factor rankings differ by vehicle type: M1 light passenger: CO > HCs > NOx > PM2.5; N freight: NOx > CO > PM2.5 > HCs. - Single M1 passenger (60–130 km/h): total emissions show a U-shape with minimum around 100 km/h (~0.28 g/km at 100 km/h), higher at 60 km/h (~0.36 g/km) and 130 km/h (~0.32 g/km). Proportional shares vary little across speeds (max ~3%). CO 0.195–0.276 g/km; HCs ~0.067–0.0685 g/km (stable); NOx 0.034–0.038 g/km (slight increase with a spike at 110 km/h); PM2.5 < 0.0025 g/km (very small sensitivity). CO decreases to 100 km/h then rises; NOx modestly increases; HCs and PM2.5 are weakly speed-sensitive. - Single N freight (40–100 km/h): CO, HCs, PM2.5 generally decrease with speed; NOx shows U-shape with minimum at 70 km/h. Ranges: NOx 1.2–1.54 g/km; CO 0.63–1.17 g/km; PM2.5 0.31–0.53 g/km; HCs 0.036–0.065 g/km. Most gaseous emissions meet limits; PM elevated (noted as affected by DPF absence stated in results). - Traffic flow (MOVES) M1: CO highest, then HCs, NOx, PM2.5 (CO roughly 100× PM2.5, ~2× HCs, 5–10× NOx). CO rate decreases with speed (rapidly at low speeds then tapering). HCs decrease then increase at higher speeds (evaporation effects). NOx decreases to a minimum at 100 km/h (~1.15 g/km reported) then rises. PM2.5 overall increases with speed but reaches a minimum around 80 km/h (~0.0067 g/km). Non-exhaust PM2.5 and PM10 vs speed show U-shaped trends; lowest at medium–high speeds (>80 km/h), with PM10 particularly reduced. - Traffic flow (MOVES) N (N1–N4): Emission magnitudes: NOx >> CO > PM2.5 >> HCs. Vehicle-type order: N4 > N3 > N2 ≈ N1 (N3 roughly double N2; N4 about double N3 for CO and PM2.5; NOx and HCs similar between N3 and N4). CO, PM2.5, HCs decrease with speed (steepest decline 80–90 km/h), then taper; NOx increases with speed (gradual to 90–100 km/h, faster above 100 km/h). Non-exhaust PM dominates PM totals (brakes/tires) and decreases with speed, with marked drops above 80–90 km/h. - Total exhaust emissions vs speed (fitted relations): • M1 single: Eps = 0.00040 v^2 − 0.00856 v + 0.71492 (R^2 = 0.927). M1 traffic: Ept = 0.00230 v^2 − 0.49410 v + 40.318 (R^2 = 0.9971). Both quadratic, minimal around 100–110 km/h; optimal low-exhaust range ~90–120 km/h, with emissions rising below ~80 and above ~120 km/h. • N single: Efs = 0.000360 v^2 − 0.05699 v + 4.44955 (R^2 = 0.8118), minimum ~70–80 km/h. N traffic: quadratic increases with speed: N1 Emit = 0.00486 v^2 − 0.60186 v + 46.6938 (R^2 = 0.9829); N2 Elit = 0.00467 v^2 − 0.57950 v + 44.94649 (R^2 = 0.9814); N3 Emet = 0.00891 v^2 − 1.07782 v + 83.39199 (R^2 = 0.9833); N4 Elat = 0.01158 v^2 − 1.60283 v + 123.51040 (R^2 = 0.9736). Growth is slow up to ~100 km/h, faster above. - Carbon emissions: • M1: quadratic vs speed with minima near 100 km/h. Single: C_ps = 0.00040 v^2 − 0.00856 v + 0.71492 (R^2 = 0.927). Traffic: C_pt = 0.00230 v^2 − 0.49414 v + 40.3180 (R^2 = 0.9971). Low-carbon range 90–110 km/h. • N: linear decrease with speed. Single: C_fs = −0.092 v + 1.3252 (R^2 = 0.978). Traffic: N1 C_mit = −0.531 v + 5.5756 (R^2 = 0.9074); N2 C_lit = −0.5335 v + 5.7472 (R^2 = 0.9109); N3 C_met = −0.5317 v + 5.5756 (R^2 = 0.9034); N4 C_lat = −1.7608 v + 19.607 (R^2 = 0.922). Pronounced reductions around 60–70 and 80–90 km/h noted. - Recommended speed ranges for sustainability goals: Light passenger (M1) 90–110 km/h; Freight (N) 80–100 km/h (balancing low exhaust and low carbon).

The findings confirm that speed is a critical lever for emission control and that optimal ranges differ by vehicle type due to engine, fuel, and load characteristics. For gasoline light passenger vehicles, both tailpipe and carbon emissions follow a U-shaped trend with speed, yielding a clear optimal band around 90–110 km/h for minimizing emissions while supporting efficient travel. For diesel freight vehicles, NOx dominates and rises with higher speeds in traffic flow, while CO, PM, and HCs decline; total emissions in traffic flow increase with speed (especially above 100 km/h). However, carbon emissions from freight decrease linearly as speed rises, indicating a trade-off between low-carbon objectives and NOx control at higher speeds. Integrating single-vehicle and traffic-level analyses reduces bias from model-specific idiosyncrasies and enables policy-relevant insights. These results support vehicle-type-specific highway speed management and variable speed limits aimed at environmental performance, complementing safety and congestion goals. The proposed ranges (M1: 90–110 km/h; N: 80–100 km/h) provide practical targets for speed limit setting and intelligent speed advisory systems to reduce both pollutant and carbon emissions.

The study integrates controlled chassis dynamometer tests with a localized MOVES micro-scale model to quantify how highway speed affects emissions for light passenger (M1) and freight (N1–N4) vehicles. It establishes that: (1) M1 emissions (tailpipe and carbon) vary quadratically with speed, minimizing near 100–110 km/h; (2) N traffic emissions increase with speed (notably above 100 km/h), while freight carbon emissions decrease linearly with speed; (3) pollutant dominance differs by type (M1: CO; N: NOx). Based on these findings, recommended optimal speed ranges to support sustainable highway transport are 90–110 km/h for light passenger vehicles and 80–100 km/h for freight vehicles, balancing low tailpipe and low carbon objectives. The work offers actionable guidance for environmentally informed speed limit policies and intelligent speed management. Future research should expand to low-speed urban contexts, incorporate evolving fleet mixes (including growing EV penetration), and refine localization/calibration to additional regions and vehicle technologies.

Single-vehicle tests used representative models and laboratory conditions, which may introduce bias due to vehicle-specific performance, mechanical wear, and controlled settings. There is an internal inconsistency noted between the listed freight vehicle after-treatment (SCR + DPF) and observed high PM attributed to DPF absence in results, which may affect PM findings. MOVES localization relied on proxy meteorology/geography (Fulton County, GA) and a model year selection (2009) to align with Chinese standards; residual mismatches may affect absolute levels. Traffic composition and operating conditions vary by road and over time; results reflect surveyed expressways and may not generalize to all corridors. Non-exhaust PM estimates carry uncertainties. Increasing EV shares and technology changes will alter fleet emissions. The study focuses on highway speeds; urban low-speed regimes (30–50 km/h) require dedicated analysis.