A 2008 Chevy Avalanche (Figure 1), with a 5.3 liter automatic engine, was setting codes P0172 and P0175 (System rich, banks 1 and 2). This vehicle, which had 94,000 on the odometer, had this problem for well over a year according to the vehicle owner and there were times the spark plugs would foul to a point where the engine would experience an engine misfire. The fix at that point was to pull the spark plugs, replace them, clear the codes and head down the road for another few months. The shop had already replaced the MAF sensor and a suspected #1 injector thinking the rich condition was being caused from a shift in MAF calculations or a leaky injector on the #1 cylinder. This however did not resolve the customer's ongoing issue.
The owner of the vehicle was very anxious to get the car resolved especially with an overdue inspection sticker on the windshield. The truck was not running well, and he thought he now had cylinder head problems because of all the carbon buildup from the prolonged rich operation and add to the fact that many people had told him that these trucks were known for bad valve guides. All the avenues had been exhausted trying to resolve his problem so he convinced the garage to perform a valve job on his truck. The garage was against it but the customer was persistent and willing to foot the bill so the cylinder heads were pulled to have a valve job done. The heads were visually examined and the shop was able to see the carbon buildup on the #1 cylinder but the other 3 cylinders on the same head did not seem bad (Figure 2). The spark plugs that came out of the engine did have black carbon on them but the #1 cylinder showed heavier deposits. The shop sent the cylinder heads out for repair and by later in the week the truck was reassembled and released to the customer.
Well, that didn’t work.
The valve job did not resolve the issue and the customer was soon back within a week with the CEL lamp on again. At this point I was now thrown into the mix to unravel the shop's dilemma. When I arrived at the shop I hooked up my scan tool to check the fuel trims with the engine fully warmed up. This would give me an indication of what the ECM was actually doing to control the fuel. The engine was not running smooth and it had a slight rough idle. The Long Term Fuel Trims were down around negative 26 and yet both the upper O2 sensors were still seeing a rich operation (Figure 3). The ECM was having a hard time in acheiving O2 switching. The whole concept of fuel control is to achieve a constant high to low switching state above and below 500 mv. The Short Term Fuel Trim will constantly move inversely l to the O2 sensor. When the O2 goes low the Short Term will go positive. When the O2 goes high the Short Term will go negative. The Short Term will eventually drive the Long Term Fuel Trim slowly to a final acceptable value of plus/minus 5 percent to 10 percent. Once the Long Term trims exceeds plus/minus 15 percent the ECM will usually set a code on most systems. I pulled the brake booster line to see if the O2 sensors would respond to a major lean condition to rule out O2 sensor failure and their readings did go below 500 millivolts. At this point I was convinced that excess fuel was coming from somewhere or the truck had a sensor that was out of range.
I next viewed some basic raw data with the engine running such as coolant and air temperature, Mass Air Flow, Throttle Position, Accelerator Position, Manifold and Barometric Pressures to see if everything was reading within reason (Figure 4). At idle, I typically expect 1 gram per liter of engine displacement per second for the MAF reading. This was a 5.3L engine, so the 7 g/sec at idle was not too far off and the other sensors all seemed to be within spec. I was looking for something that could really add some fuel such as leaky injectors, high fuel pressure or a leaky purge valve. So I first disconnected the purge valve line at the left side of the engine just above the ignition coils (Figure 5) but this did not have any effect on the rich operation. I even placed a fuel pressure gauge on the rail and measured the pressure, finding it within the spec of 60P.S.I. and when I turned the engine off the fuel pressure held without any signs of leakdown.