风机设备

Alright, let's finalize the ID fan selection. We're specifying two 50% capacity fans for the 660 MW unit — normally both running, each handling about 55% of the design flow under normal conditions. Francesco, what are you offering?

We've proposed our ACF-4000 series — double-inlet, backward- curved centrifugal fans with inlet vane control. Impeller diameter is 4.2 meters, design speed 745 rpm via a 6-pole motor direct drive. At the design point, each fan delivers 620 m³/s at 5,800 Pa total pressure, with a fan static efficiency of 86.5%.

Francesco, the 86.5% efficiency — is that at the guaranteed point or the test block average? Per ASME PTC 11, we need the average across the operating range, not just the peak.

That's a fair question. 86.5% is at the design point. Across the operating range from 70% to 100% flow, the weighted average static efficiency is 84.2%. We guarantee a minimum of 83% average. We've also included a ±2% efficiency tolerance band per ASME PTC 11.

I'm more concerned about part-load operation. During low-load periods — say 40% boiler load at night — each fan would be pushing about 250 m³/s. That's only 40% of design flow. With inlet vane control, what's the efficiency at that point?

(pulls up part-load curve) At 40% flow with inlet vane control, fan efficiency drops to about 62%. The alternative is VFD — variable frequency drive. With VFD, at 40% flow, efficiency would be around 78%. The VFD option adds about $280,000 per fan, but the energy savings could pay that back in 2-3 years depending on the load profile.

That's significant. Hannah, what's the projected annual operating profile? How many hours at part load?

Based on the grid dispatch forecast, we expect about 2,800 hours per year at 40-60% load, especially during the winter months. The rest of the year — about 5,200 hours — at 80-100% load. So roughly 35% of the time at part load.

Let me run a quick payback calculation... At 40% load, the VFD saves about 16 percentage points of efficiency. That's roughly 250 kW per fan × 2 fans × 2,800 hours × $0.08/kWh = $112,000 savings per year. The $560,000 incremental cost pays back in exactly 5 years. Borderline.

Don't forget the maintenance benefit. With VFD, you eliminate the inlet vanes — fewer moving parts, less wear. And soft- start reduces mechanical stress on the coupling and shaft. It's not just about energy.

Good point. Let's go with VFD then — it's the right long- term decision. Francesco, update the proposal with VFD drives and confirm the delivery schedule. Wang Jian, you'll lead the performance test planning?

Yes. We'll follow ASME PTC 11 for the field performance test. I want to witness the factory performance test too — the AMCA 210 test in their lab. Verify the efficiency curve before the fan ships.

One more thing — noise. These are massive fans, and the plant is only 800 meters from a residential area. What's the predicted sound pressure level at the site boundary?

Without attenuation, 98 dBA at 1 meter from the fan casing. We're including an acoustic enclosure that drops it to 82 dBA at 1 meter. With distance attenuation to 800 meters, the predicted noise at the boundary is 38 dBA — well below the 45 dBA nighttime limit.

Satisfied. Let's proceed. Good technical review, everyone.

Carlos, the cooling tower fan CTF-3B tripped on high vibration — 12.5 mm/s on the fan bearing, alarm set at 7.1 mm/s. We tried restarting but it trips again within two minutes. We need this fan back — we're losing cooling capacity.

That's a serious spike. Zhou Lei, grab your vibration analyzer. Ivan, lock out the fan. Let's get to the top of the cooling tower and figure out what's going on.

(at the cooling tower top, fan locked out)

(mounting accelerometers on the bearing housing) OK, let me take baseline readings with the fan off first. Ambient vibration... negligible. Now, can we do a slow roll?

Ivan, start the fan at 10% speed — jog mode. Just enough for one revolution.

Jogging... done.

OK, I'm seeing a strong 1× running speed component — that tells us it's an unbalance issue, not misalignment or looseness. If it were misalignment, we'd see a strong 2× peak. If it were looseness, we'd see harmonics all the way up. Unbalance is actually the best-case scenario — we can fix it with in-situ balancing.

In-situ balancing on a cooling tower fan? We've never done that here.

No time like the present. Here's the principle: we measure the vibration amplitude and phase angle at running speed. Then we add a trial weight at a known position, run again, measure the change. The change tells the analyzer how much weight to add and exactly where to place it. The whole process takes three or four runs — about an hour.

What caused the unbalance? These fans have been running fine for two years.

(inspecting the blades with a flashlight) There — look at blade number 4. Scale buildup on the leading edge, about 3 mm thick. The other blades are clean. That scale is from the cooling tower water mist — it's basically mineral deposit. That asymmetry is your unbalance.

So we should clean the blade instead of adding weights?

Cleaning would certainly help, but it's hard to clean perfectly in the field. Plus, the scale will come back. Better approach: we balance it now by adding weights, and schedule a full blade cleaning during the next planned outage. That way we get the fan running today and fix the root cause later.

Agreed. Zhou Lei, let's do this. Ivan, prepare the balance weights — we'll need the bolt-on type for the hub. Nalini, I'll keep you posted on progress.

(after four balancing runs)

Final run: vibration at the fan bearing is 2.8 mm/s — well below the 4.5 mm/s alert and miles below the 7.1 mm/s alarm. Phase angle is stable. Fan is balanced. Carlos, sign off on the balancing report.

Beautiful work, Zhou Lei. First in-situ fan balancing at this plant — took us 55 minutes from lockout to sign-off. Nalini, CTF-3B is cleared for normal operation.

Excellent. I'll restart now. And Carlos — write up a work order for blade cleaning on the next outage. We don't want to do this balancing act on all three fans!