EPA 2027 NOx Rule Expected Late June — 35 mg/hp-hr Limit Stands

What NOx limit will 2027 engines have to meet?

The EPA is expected to finalize a 35-milligram-per-horsepower-hour NOx emissions ceiling for heavy-duty truck engines by late June, according to OEM statements. That's an 82.5% reduction from the current 200 mg/hp-hr standard that has been in place since 2010.

Engine manufacturers told Transport Topics they expect the agency to leave the 35 mg/hp-hr threshold unchanged from the draft rule published earlier. The final rule will dictate the aftertreatment hardware, calibration complexity, and service intervals for every Class 8 tractor built for model year 2027 and beyond.

What the 35 mg/hp-hr standard means for aftertreatment hardware

Meeting 35 mg/hp-hr will require larger selective catalytic reduction (SCR) systems, additional diesel exhaust fluid (DEF) capacity, and more frequent regeneration cycles. Cummins, PACCAR, and Navistar have all confirmed they are engineering new aftertreatment packages to hit the standard, though none have released final packaging dimensions or weight penalties yet.

The tighter NOx limit also means more sensors, more software calibration updates, and a narrower margin for sensor drift before a fault code triggers. Shops that service 2027-and-later engines will need scan tools capable of reading expanded diagnostic trouble codes and recalibrating NOx sensors to tighter tolerances than current X15 or MX-13 platforms require.

DEF consumption and tank sizing

Fleets should expect DEF consumption to climb. Current engines use roughly 2–3% of diesel fuel volume in DEF; the 35 mg/hp-hr standard will likely push that ratio closer to 4–5%, based on preliminary OEM guidance. That means a truck running 120,000 miles per year on 6 mpg will burn an additional 200–400 gallons of DEF annually compared to a 2026 model.

Tank sizing becomes critical. If OEMs add DEF capacity without expanding the frame rails, diesel tank volume may shrink, cutting range. If they don't add capacity, DEF fill intervals drop below 5,000 miles for some duty cycles — a service-stop penalty for long-haul operators.

Service interval and warranty unknowns

No OEM has published 2027 service intervals yet. The added aftertreatment complexity — more substrate volume, tighter sensor tolerances, higher regeneration frequency — historically correlates with shorter oil-change intervals and more frequent DPF cleanings. Fleets running 2027 engines should budget for at least one additional service stop per 100,000 miles until real-world data proves otherwise.

Warranty coverage for the new NOx hardware is still undefined. The EPA's emissions-defect warranty runs five years or 100,000 miles for heavy-duty engines, but that covers only emissions-related failures. If a larger SCR canister cracks due to road debris or a NOx sensor fails outside the defect window, replacement cost falls on the fleet. Aftermarket pricing for 2027 components won't be visible until parts hit distribution in late 2026.

What fleets should do before the rule drops

Order 2026 model-year inventory now if your replacement cycle allows it. The 200 mg/hp-hr engines are proven, parts are stocked, and technicians know the failure modes. The 35 mg/hp-hr hardware will be unproven at launch — no 500,000-mile durability data, no field recall history, no aftermarket cross-reference yet.

For fleets that must spec 2027 units, plan for higher upfront cost (OEMs have not disclosed pricing but aftertreatment hardware alone will add $3,000–$5,000 per truck based on supplier estimates), reduced payload from heavier emissions systems, and a learning curve for your shop. Budget an extra $0.02–$0.03 per mile for DEF and assume service intervals will tighten until OEMs prove otherwise with oil-analysis data.

The June finalization timeline means OEMs have roughly four months to lock production tooling and supplier contracts before 2027 model-year build starts in October. Any last-minute changes to the 35 mg/hp-hr standard after June would force delays or force manufacturers to over-engineer systems to cover a range of possible limits — both scenarios that increase cost and risk for the first year of production.