High CO and Low O2 After Annual Burner Service
A 500 hp firetube boiler burning natural gas had its annual burner service last week. Since then the stack shows a light blue-gray haze. A portable analyzer at the stack sample port reads O2 1.8%, CO 900 ppm, and stack temperature 495°F. Last year's commissioning sheet, at the same firing rate, shows O2 4.2%, CO 15 ppm, and 405°F. Overfire draft is −0.02 in. w.c., matching the baseline. Burner gas manifold pressure is 3.9 in. w.c. against a nameplate 3.5 in. w.c. at high fire. The boiler is at high fire and steady. The maintenance lead wants to open the burner air shutter 'two turns' and watch the stack clear.
1. Interpret the readings. What is the most likely story, and what does the 3.9 in. w.c. manifold pressure tell you?
Read the worked answer
The burner is running rich - it is getting more fuel than it has air to burn.
- O2 1.8% vs. 4.2% baseline: the excess-air margin has collapsed. For natural gas you normally want roughly 3–5% O2 at high fire (about 15–25% excess air), depending on the burner's commissioning data.
- CO 900 ppm vs. 15 ppm: incomplete combustion. CO is a direct sign that fuel is being broken down but not fully oxidized - either not enough oxygen, or the flame is being quenched below the temperature needed to finish the reaction.
- Stack temperature 495°F vs. 405°F: roughly 90°F of extra heat is going up the stack. Some of it is unburned fuel and some is simply a larger flame releasing more heat than the tube bundle can absorb. It is also a real efficiency and thermal-stress issue on the fireside.
- Manifold pressure 3.9 in. w.c. vs. 3.5 nameplate at high fire: this is the key clue. The burner is being over-fired relative to its design. More gas at the same air setting equals a rich mixture.
- Overfire draft unchanged at −0.02 in. w.c.: draft is not the obvious culprit, so look at the gas train, the burner internals, and the combustion air path.
Do not overlook the health hazard: 900 ppm CO in a flue does not stay in the flue. Get a CO meter into the breathing zone immediately and ventilate.
2. In what order do you check, and why is opening the air shutter 'two turns' the wrong first move?
Read the worked answer
Why not just open the shutter: it is a guess, and a guess that can make the real problem harder to find. If the root cause is a gas regulator passing high, opening the shutter masks it; later, when gas pressure drops back toward normal, your linkage is now in a position that does not match the commissioned curve and the burner goes lean and unstable. A burner is tuned by measurement against the commissioning sheet, not by watching the stack clear.
Order of work:
- Safety first. CO monitor in the room, ventilate, notify the plant, have your hand on the stop. A boiler producing 900 ppm CO at the stack is one that may be producing CO in the boiler room.
- Verify the instrument. Calibrate or bump the analyzer with span gas, and sample at the same port, upstream of any air in-leakage, as the baseline readings. A bad cell or a leaking sample line has fooled many people.
- Confirm steady state at high fire - at least 10–15 minutes at the same firing rate as the baseline. Do not tune against a boiler that is modulating.
- Check the gas side first. Manifold pressure at high fire against nameplate (3.9 vs. 3.5 is out of spec). Look at the regulator, the spring, lockup, and whether the gas orifice or metering plate was changed or left out during the service.
- Inspect the burner internals. Diffuser position, gas spud orientation, orifice size, gaskets, any part left out, reversed, or installed in the wrong clock position after the annual service. Post-service assembly errors are the single most common cause of a sudden combustion problem.
- Then the air side: air shutter and linkage reference marks against the commissioning sheet, FD fan, inlet louvers, filters, combustion air openings, and any flue gas recirculation.
- Tune only after the cause is found: low fire first, then high fire, then re-check low fire, because CO is usually worst at low fire.
Document the before and after numbers in the log.
3. Suppose high fire cleans up nicely but low fire still shows 400 ppm CO with visible O2. What is happening, and what do you do?
Read the worked answer
This is flame quench, not a fuel/air ratio problem at high fire.
When CO persists at low fire while high fire is clean, the flame is being cooled below the temperature at which CO finishes burning to CO2. Common causes:
- Excessive excess air at low fire - too much cold air cools the flame. Counterintuitive but very common: more O2, more CO.
- Low-fire gas pressure too high for the low-fire air - the low-fire position was set up wrong, or the modulating motor's low-fire cam or linkage slipped during service.
- Flame impingement or flame too close to a cool surface - burner insertion depth, diffuser position, or a cold furnace at light-off. Some CO at cold start that clears as the furnace warms is normal; CO that persists at steady low fire is not.
- Burner repositioned during the service so the flame pattern is wrong.
What to do: check the low-fire gas pressure and air damper against the burner manufacturer's low-fire settings, check the low-fire cam and positioning motor, check burner insertion. Re-tune and retest at steady low fire. If you cannot get the CO within the burner's commissioning spec and the acceptable limit, keep that boiler off line and run the other boiler. Do not accept 'small' CO numbers on a boiler you have to live with - persistent CO is telling you the burner is not set up correctly, and it will get worse, not better. Notify the burner representative and get it in writing.