Pump Danger

By Josiah Wagener

[Josiah Wagener recently read the Lost Horse Creek Fire Pump Burn Injury RLS in the Wildland Fire Lessons Learned Center’s (LLC) Incident Review Database (IRDB). This scalding burn injury reminded Josiah of a similar incident that happened to him in 2003. It’s never too late to share your lessons!]

I just read the Lost Horse Creek Fire Pump Burn Injury RLS about a scalding injury from the suction hose blowing off of a Mark-3 pump. I had a similar incident in 2003. Fortunately, there were no injuries. At the time, I understood why it had happened, but it still seemed to me like an utterly freak occurrence.

However, after looking through the various reports in the LLC’s IRDB, I now I think this type of pump incident may not be quite as rare as I had assumed.

Image of the pump and burn injury from the Lost Horse Creek Fire incident RLS.

Remembering the Story

I was running the pumps for my crew on a fire in Montana. I made the mistake of thinking I should set up parallel rather than series pump and hose lay to get more pressure up a steep mountainside.

In hindsight, I know that was a mistake, but it had been a several years since I had run a complex pump system and I was a little fuzzy on the basic hydraulic rules. I had two Mark-3 pumps next to a porta-pond with a few hundred feet of hose running uphill from each pump before joining together with a reversed wye valve and continuing several hundred feet farther up the hill in a single hose.

I had check/bleeder valves on both of the primary hose lines. I believe that played a part in what went wrong. The mountain was pretty high in elevation, so there was a lot of head pressure pushing back against the pumps.

I had both pumps running at full power. I think one of them must have been producing a little more pressure than the other. That meant that the pressure all loaded against the check valve on the weaker pump. In retrospect, I think that pump ended up moving no water at all for the first hour or so of work. As far as I can recall, it did take about an hour of full power operation before the trouble manifested.

I had tried to check the pressure by feeling the hoses, but of course they would both have been rock hard—so that wouldn’t really tell me anything. I don’t remember what the first sign of trouble was. I may not have realized that anything was wrong until the suction hose blew off in a cloud of steam. This all happened 20 years ago. While I don’t remember all the details, I sure do remember that suction hose blowing off!

Dramatic Incident

Fortunately, there was nobody right next to the suction hose when this happened and the amount of actual boiling water in the pump was relatively small. In fact, while I remember it as a pretty dramatic incident, I don’t think I had ever really thought about how dangerous that steam might have been until I read the recent report on the Lost Horse Creek pump scalding injury incident.

I’m sure that the pump spinning for an hour without moving any significant amount of water was what caused it to boil, but I think that the suction hose actually blowing off of the pump probably happened because the check valve on that pump broke, allowing all of the head pressure to slam back down through the pump. I believe that the check valve had been vibrating under the competing water pressures that entire time and it finally failed catastrophically. After the accident I examined the check/bleeder valve, along with all of the other hardware, and discovered that the clapper in the valve had, in fact, shattered.

Once the suction hose with its foot valve/strainer was gone, there was no check on the water and all the water in the hose lay began flowing backward through the pump. That caused another rather strange hazard because I couldn’t get the pump to shut off.

If I recall correctly, it would cut out for a couple seconds every time I pushed the kill button, but then it would fire back up again as soon as the kill button was released because the backflow of water was constantly jumpstarting the motor.

The motor must have been turning backwards as the water would have been pushing the pump impellers backwards. I don’t really understand how it could run that way, but it did. After a couple unsuccessful attempts to stop the pump, I ran a few hundred feet uphill to the reversed wye valve where the hoses joined together and shut off the valve to stop the head pressure flowing back into the pump. That finally allowed it to shut down.

Once I had it shut off, I noticed that the paint had begun to blister a little bit on the pump head due to the heat. In hindsight, if I had been touching the pump heads periodically I might have noticed the temperature rising–but I didn’t think to do that at the time. I trusted that as long as the guys at the nozzles way up atop the hill were getting water, the pumps must be working properly.

With the pump shut down I managed to strip the pump head off of the motor of the Mark-3 and tossed it into the porta-pond to cool off, then began identifying and flagging broken parts in the system. The other pump, the one which I suspect was slightly stronger of the two, had continued to run throughout the entire incident. The guys using the nozzles up at the top of the hose lay hardly noticed anything. They hadn’t had very much water pressure to begin with because they were more than 800 feet uphill. They only lost water pressure for a minute or two before I shut off the effected side of the wye.

Lessons

To me, the lesson from this event is if you don’t do complex pump and hose lays on a regular basis, you should review the NWCG S-211 Portable Pumps and Water Usage class every few years. I had taken S-211 about six years before this incident, but only maybe twice during those years had I done anything more complicated than running a small progressive hose lay off of a single pump.

On the other hand, neither my Superintendent nor my Squad Boss—nor anyone else on the crew—pointed out that setting the system up in parallel rather than series was the wrong way to do it.

Presumably, none of them recognized the mistake any more than I did. Most of them had also probably taken S-211 at some point. However, I have noticed many firefighters’ eyes tend to glaze over in class when learning hydraulic principles and hydraulic calculations. S-211 seems to be one of those often discounted classes. You take it once and you’re done because, after all, portable pumps are so simple anyone can run them, right? Or maybe we just assume that all firefighters run pumps so often that we don’t need refreshers. Personally, I suggest that pumps and complex hose lay refreshers should be done regularly every few years.

I have never set up another parallel pump operation since then, And I wouldn’t do so again if I had anything more than a trivial amount of head pressure to work against. It is impossible to know the exact pressure coming out of a portable pump. Therefore it’s all too easy for one pump to overpower the other pump—as was my experience on that fire in Montana in 2003.

The recent Lost Horse Creek scalding injury case didn’t involve parallel pumping. But it did involve two pumps working against each other and also some pretty big head pressure leading to a pump which was moving effectively no water for a significant amount of time. I would be interested to know if their lower pump had a check/bleeder valve on its hose. I don’t think that was specified in the report. I suspect that if there were no check bleeder valve then, even against high head pressure, that pump might have been able to get at least a little bit of water to flow. But with a check valve in place the pump has to be able to completely overcome the head pressure in order to move any water at all. Of course we need to have some form of check/bleeder valve in order to get a pump restarted when under head pressure. However, I suspect they also contribute to the problem in rare events such as these.

In searching through the LLC’s IRDB I found another similar incident on the Crescent Fire in California in 2017. In that case two U.S. Forest Service Type 3 engines set up a parallel pumping system up a steep hill. In that incident, similar to my incident, the two pumps were competing and the head pressure meant that one pump was moving effectively no water. The pumps on the engines had bypass valves which were supposed to allow a little water to flow back into the engine tank under those conditions, but one of those bypass valves failed to operate and a similar steam explosion occurred–once again leading to a firefighter badly scalded by boiling water.

How Many Near Misses?

I wonder how many near misses there have been due to this type of problem that no one knows about because no one was injured?