Exhaust gas recirculation (EGR) is a complicated emission control technology that’s bad for fuel economy, performance and maintenance for diesel engines. This post explains why EGR is bad and compares three DIY modifications to stop EGR: 1) unplug, 2) “blanking plate”, and 3) “EGR delete harness” (which fakes the intake air temperature). The EGR delete harness seemed to have no bad side effects.
Disclaimer: Disabling emission control systems is illegal for road use. Modifying your vehicle is your responsibility. What works on my vehicle may not work on your vehicle.
EGR is bad
EGR is an nitrogen oxides (NOx) emission control technology. It’s an afterthought, to meet increasingly onerous emissions standards. If you’re serious about reducing emissions (and saving money, reducing traffic congestion and making our lives more peaceful and healthier) then think about how you can drive less!
There are many reasons why EGR is bad for diesel engines:
- EGR deliberately reduces combustion temperatures to prevent NOx emissions. Lower combustion temperatures reduce combustion efficiency, reduce power and increase soot (and black smoke). Diesel particulate filters are yet another troublesome pollution control technology to reduce soot emissions.
- EGR gases are not simply carbon dioxide and water vapour (resulting from the oxidisation of hydrocarbons) and nitrogen (which shoudln’t be oxidised). Recirculated soot combines with oil mist from the positive crankcase ventilation system (that draws blowby gases into the air intake) to deposit in the intake manifold, choke the engine and clog EGR valves and swirl control flaps (if present). On cold start, the cold engine block acts as a large heat sink that reduces combustion temperatures and combustion efficiency, resulting in unburnt fuel (which is why diesels have glow plugs that attempt to warm up the combustion chambers). EGR should not be activated until the engine warms up, however it is active at cold start on some vehicles. Recirculated fuel will result in a surplus of fuel, unevenly distributed among the combustion chambers, making the engine produce loud “knocking” noises and blow white smoke.
- EGR gases are hot. Intake air is preferably cold, which is why turbocharged diesel engines have an intercooler. Cold, compressed air is dense, contains more oxygen and can burn large volumes of fuel. Mixing of EGR gases with intake air increases the temperature and somewhat defeats the purpose of the turbo intercooler.
- EGR needs to be carefully controlled. Too much EGR and the engine runs bad. Too little EGR and NOx emissions increase. In turbocharged engines, the pressure differential between the exhaust manifold (exhaust backpressure) and intake manifold (intake pressure) needs to be carefully controlled. This can be achieved with even more technology such as a variable position EGR valve, variable geometry turbo, manifold air pressure and temperature (MAP) sensor and an intake shutter flap.
There is sufficient real world evidence that EGR is bad:
- EGR is stopped during acceleration because EGR robs the engine of power. Volkswagen went one step further and stopped EGR altogether for road driving to increase fuel economy (and thought nobody would discover that real world NOx emissions were up to 40 times greater than those during laboratory testing).
- Big trucks don’t have EGR because it’s bad for fuel economy and maintenance. Big trucks use selective catalytic reduction technology, where a reductant such as aqueous urea (diesel exhaust fluid) is injected into the exhaust to convert NOx into harmless nitrogen and water vapour. Consumer diesels typically do not use SCR technology because of the cost, the inconvenience and, let’s be honest, modern consumer vehicles are not designed to be high mileage, long-life vehicles.
There are basically two DIY strategies for disabling EGR: 1) physically disable the EGR valve, and 2) fool the engine control unit (ECU) to keep the EGR closed (a “EGR defeat” hack).
The EGR can be physically disabled by unplugging the electrical connector that controls the EGR opening/closing (for example) or by installing a “blanking plate” in the EGR pipe to block the flow of exhaust gases. Some blanking plates have a hole just large enough to prevent check engine lights (CEL).
In many vehicles, the ECU closes the EGR valve when the ambient air temperature (AAT), measured at the mass air flow (MAF) sensor is cold. I suppose the reason is that cold air is sufficient to reduce combustion temperatures and NOx emissions. The AAT sensor is typically a resistance temperature device. By increasing the resistance, a fake low AAT can be achieved. The cheapest way is to solder a resistor onto the harness, in parallel to the sensor. You can also buy plug in harness extensions that pass through the MAF sensor power and signal, isolate the AAT sensor and include a resistor to fake the AAT resistance. These “EGR delete harnesses” are popular because they are easy to install and can be quickly removed before servicing or emissions testing.
Whether or not AAT is part of the ECU “mapping” that controls the engine has been a subject of much debate. In theory, temperature affects air density and must affect mass air flow rate. Using the ideal gas law, the MAF error for 50 °C real temperature and 0 °C fake temperature is about 18%, i.e. the fake MAF is 18% larger than the real MAF. However, the service manual for my vehicle shows a simple graph relating MAF in grams/sec to the MAF sensor hot wire voltage with no mention of AAT. Furthermore, the turbo will compress the intake air, the intercooler will cool the air some and then it is introduced into a warm intake manifold, where there is a MAP sensor to measure pressure and temperature.
An inexpensive ELM327 “scan tool” that can read standard on-board diagnostics parameter IDs (PIDs) is sufficient for measuring the effect of various EGR modifications. I used an OBDLink LX bluetooth ELM327 adapter and the FORScan Lite app on my Android phone. FORScan is excellent for Fords and older Mazdas because it also reads the manufacturer PIDs that mechanics want.
I tested five EGR modifications on my Mazda BT-50 3.0L diesel: 1) none, 2) “holy” blanking plate, 3) blanking plate, 4) EGR control solenoid unplugged, and 5) EGR delete harness. I installed the blanking plates at the upstream flange of the EGR cooler. This location was the easiest to access and if the EGR cooler was to leak, a solid blanking plate will stop coolant from reaching the EGR valve.
I recorded several PIDs, including: ambient air temperature (at the MAF sensor, AAT), EGR control solenoid state (EGRV2), EGR solenoid duty cycle (EGR_PCT.OBDII), intake air temperatre (at the MAP sensor, IAT.OBDII), mass air flow rate (MAF.OBDII) and intake manifold air pressure (at the MAP sensor, MAP.OBDII). I recorded these PIDs under three conditions: 1) at idle (stationary), 2) slow driving at 60 km/h, 4th gear and 3) highway cruising at 90 km/h, 5th gear (I don’t like driving fast plus there were frequent roadworks). I recorded the PIDs at least three times and have reported the median values (some measurements like MAF do vary). Below are three graphs showing the effect of the five EGR modifications.
Danger: Two people are needed for road testing: one person drives and one person records the PIDs. Do not look at your phone while driving. It’s dangerous.
The holy blanking plate was not effective. The ECU commanded an increase in EGR_PCT.OBDII, to open the EGR valve more and thereby compensate for the restriction in the EGR pipe. MAF.OBDII and IAT.OBDII changed little from the unmodified values.
The blanking plate stopped EGR flow. MAF.OBDII increased and IAT.OBDII decreased relative to the unmodified EGR. From the differences in MAF.OBDII, the unmodified intake was 59% EGR gases at idle, 54% at 60km/h and 21% at 90 km/h. The ECU didn’t know about the blanking plate and increased EGR_PCT.OBDII to attempt to increase EGR flow and decrease MAF.OBDII. After some 10s of kilometres, the ECU turned on the CEL. There were two error codes: 1) P0401-FF EGR system insufficient flow, and 2) P0401-C EGR system insufficient flow. Driving with the CEL always on is risky because it masks any new and potentially important faults.
Unplugging the EGR control solenoid forced EGRV2 off (EGR closed). MAF.OBDII and IAT.OBDII were similar to those values with the blanking plate. Clearly the EGR valve was closed and not leaking. EGR_PCT.OBDII was 5%, which is normal for this vehicle with the EGR closed (note: EGR_PCT.OBDII is the commanded EGR solenoid duty cycle and not the actual EGR valve position, for which there is no PID). Unplugging the EGR control solenoid state produced a CEL as soon as the ignition was switched on.
The EGR delete harness I purchased has a 4.7 kohm resistor across pins 1 and 2. The RTD in the MAF sensor measures about 1.3 kohm at about 25 °C. With the EGR delete harness, AAT was constant 4 degC and EGRV2 was off (EGR closed). MAF.OBDII and IAT.OBDII were similar to those values with the blanking plate and the EGR unplugged, from which I concluded that there was no substantial change in the ECU mapping with the EGR delete harness and fake AAT. There was no CEL. I have installed the blanking plate together with the EGR delete harness and still there are no CELs after hundreds of kilometres of driving.