High Fire Combustion Readings That Do Not Add Up
A 500 hp, 150 psig packaged firetube burns natural gas with forced draft, parallel-positioning fuel/air linkage, and an oxygen trim system. Commissioning data at high fire: 3.0% O2, 400°F stack temperature, 20 ppm CO. Last week the building added an enclosed combustion air intake plenum and relocated a louver section. This week's test at low fire is essentially unchanged: 3.5% O2, 395°F, 15 ppm CO. At high fire the same analyzer now reads 8.5% O2, 520°F stack, and 250 ppm CO, and the CO number is drifting upward during the test. The flame appears short and bushy with slight pulsation. Steam load is normal. Gas supply pressure is 7 in. w.c. static and holds 3 in. w.c. at high fire. The O2 trim is in automatic. Efficiency calculated from the new data is about four points worse. Nothing has been done to the burner since the plenum was installed.
1. Do these readings show a lean condition, a rich condition, or something else? Explain how 8.5% O2 and 250 ppm CO can coexist.
Read the worked answer
Neither label fits cleanly, and that is the point of the question. Oxygen at the stack tells you how much air passed through the furnace unused. CO tells you how much fuel failed to find oxygen where the flame was. They are measured at different places in the process, and they can both be high at once when the problem is mixing rather than quantity.
8.5% O2 means a large surplus of air is leaving the boiler. 250 ppm CO with a bushy, pulsing flame means part of the fuel is being quenched or lifted off the diffuser before it can burn - the air is there, but it is not reaching the fuel at the right velocity and turbulence. Excess air that blows the flame off the burner head is a classic cause of exactly this pairing, and it also raises stack temperature, because you are heating a larger mass of air and dumping it up the stack. That accounts for the 120°F rise and the four efficiency points.
The wrong response is to trim fuel up until CO drops. That would restore a ratio on paper while leaving the flame unstable, and it risks a fuel-rich condition the moment load changes. The right framing is: find what changed on the air side, restore stable combustion first, then set the ratio. Low fire being unchanged is consistent with a fault that only appears at high fire, where air velocity and damper position are greatest.
2. Give your diagnostic sequence. What do you check first and why?
Read the worked answer
Start with the only thing that changed: the air path. A new plenum and relocated louvers alter both the available free area and the pressure drop upstream of the forced draft fan. Check louver free area against the burner's required cfm, measure the negative pressure in the plenum at high fire, and inspect for a damper, screen, or construction debris left in the new ductwork. Then watch the air damper itself at high fire while someone drives the firing rate: confirm the actuator reaches its full-open position, that the linkage set screws are tight, that the damper blade actually moves with the shaft, and that the positioner feedback matches the blade. A slipped linkage on the air side produces exactly this signature at high fire only.
Next, verify the fan can deliver: belt condition and tension, fan amps compared with commissioning, and whether the fan is running near the end of its curve. Then look at the burner itself - diffuser and air swirl plates for dirt or distortion, gas spuds for obstructions, and flame shape at each firing rate rather than only at high fire. Confirm gas pressure holds at the burner manifold under load, which it does here.
Calibrate the analyzer before trusting any of these numbers. Sample the stack with a warmed, filtered probe and check for air in-leakage at the sample port. A false-high O2 reading would make the trim system respond to a condition that does not exist, and you would chase the wrong problem for a week.
3. Repair cannot be completed before the night shift. What operating limits and monitoring would you set, and what would make you shut the boiler down?
Read the worked answer
With a combustion problem that is not fully understood, the safe operating envelope is narrow. Limit firing rate to the highest point at which the flame is stable and CO stays low and steady - in practice that usually means capping at mid fire and taking the load elsewhere or accepting reduced steam capacity. Lock or disconnect the O2 trim so it cannot keep driving the air damper while the fault is unresolved; manual positioning with a documented setting is more predictable than an automatic loop chasing a bad signal. Watch the flame through the sight port each round and record what you see, not just what the meter says.
Monitoring: a portable CO analyzer at the stack and a CO monitor in the boiler room, readings logged hourly along with stack temperature, O2, and firing rate. Alert the building on the CO monitor. Confirm the boiler room ventilation and the low-water and flame-safeguard interlocks are functional before leaving the machine on line.
Shut down for: CO rising above the alarm level or climbing without a load change, visible flame instability or rumble, flame loss, a burner trip that will not relight cleanly, or any sign of a fuel-rich condition. Do not leave a machine running on the assumption that the trim will sort it out; trim systems have no way to tell the difference between a lean furnace and a lying probe.
After repair, re-establish the baseline at low, mid, and high fire with a calibrated analyzer, then re-enable trim and document the new numbers against the commissioning sheet.