Upsets

OK, it's time to displace that post about the worn out tire :)

I've been working too much and not getting enough days off. Meanwhile I've been taking care of some tasks that are required to be completed before year's end. They are alas turning out to be a bit more involved than I'd realized. No matter, I'll get them done.

We managed to have a couple of nasty upsets yesterday. Once in the morning before getting off work, and another shortly after we'd arrived for the following night shift.

The first event was a bit weird. The first inkling we had that something was amiss was a light rumbling noise, which got louder and louder, until it was shaking the windows of the control room. I got up and poked my head out the door to see if it was coming from the facility next door (natural gas flaring off, or something), but it was coming from our plant.

We quickly scanned through the screens and learned that the high pressure steam vent valve had come open 100%. We also found that it could not be placed in manual and shut from the control room. I grabbed my jacket and earplugs and headed for the top of the boiler (about 7 stories up).

Approaching a 1200 psi steam line roaring to atmosphere is difficult to describe. Not many people would care to get near such a release of energy. Every cell in your body vibrates from the noise. The shattering chest-rattling thunder of Niagara falls, concentrated down to a raging hot valve you have to shut off. So I did. This wasn't the first time I've had to do this - just the first time recently and the first time I've actually written of it.

I figured that we'd lost quite a bit of level out of the boiler, with the vent open so long, and sure enough, we had. As I made my way down off the boiler I heard the feedwater pumps starting and stopping... mind you these are big pumps, and you don't get many starts before the motor protection relays decide the windings are hot enough and prevent further starts.

So I got back to the control room and sure enough, the boiler was low on level and they were trying to keep a feedwater pump going. Problem is when you really need the water and open up the feedwater valves, the pumps lose suction pressure and trip off on NPSH. So they'd had a few issues trying to fill the boiler too quickly while I was on the boiler. We cut a little load off each turbine (to cut down on steam production) and slowly started building levels again. Eventually everything got back to normal and we were cruising with everything back in automatic again.

The cause of all this fuss? One steam pressure transmitter with an output signal that suddenly started spiking to a couple of thousand psi, telling the control system to vent the pressure off. I think we need some redundancy here...

This shift's moment of truth: Loss of fuel gas.

We'd been here perhaps two hours when suddenly the alarm for low fuel gas pressure came in. I pulled up a real-time trend, and pressure was plunging off the cliff. Both gas turbines went into a runback - this reduces the rate of pressure loss and also reduces the odds of a flame-out at full load, which is an ugly thermal stress on the machine. The issue of thermal stress cannot be overstated - many of the metal parts in the combustion path are ceramic-coated, and rapid cooling of the metal causes cracking and loss of the ceramic coating that protects the metal parts from hot corrosive exhaust gas.

I called the foreman's office for our gas supplier and told him we were losing fuel pressure. Then I looked out the door to see what was up with them. All their flares were blazing - although it was night you could read a book like the sun was up - fierce shadows are cast by the bright yellow fireballs on top of the flare stacks.

Meanwhile both gas turbines had runback completely and their generators had come off the grid. The turbines stayed lit however and hung on at 3600 rpm. The steam turbine began losing load as the boiler pressure decayed. With the gas turbines offline, there was not much exhaust gas to generate steam. Gas pressure stabilized, and then started to recover. Within a couple of minutes, fuel pressure was back to normal. We quickly synched the generators and loaded the gas turbines up. Had to wait a couple more minutes to get the fuel gas temperature back up, then up to full load and into emission compliance.

The steam turbine, being quite a laggard, didn't start recovering until both gas turbines were at full load for a few minutes. I was very happy we didn't lose either gas turbine, because when they flame out, you are stuck with a 15 minute boiler purge (for each one!) to eliminate any combustibles. And we would have lost the steam turbine too. Talk about a slow recovery time. At least this way it was over with in 15 minutes.