Fire Pump Suction Minium Tank Level

[Smokerr]

I am in the process of setting a procedure for dealing with whats probably a fairly typical situation but I can find no information on it in regards to what occurs at the low end of the tank water levels.

We have a 29 foot supply in the tank with an intake thats 3 feet with an Anti Vortex plate on top of that. Somewhere around 3 feet we will of course loose suction. The question is how far above 3 feet would that occur (I am currently using 5 feet as the critical point as at full flow that allows about a minute worth of options)

Basis for the Concern:

Our fire pumps use heat exchanger and the cooling is per a discharge tap off the supply side.

Once the discharge pressure goes away, the pumps will self destruct due to lack of cooling, i.e. the engines will seize and then things will start to break. I suspect engine will come apart and possibly couplers and a lot of inertia back into the engine from the pump rotating assembly (3500 gpm pumps at 100% so its a pretty large mass)

Despite redundant pumps no provisions were made to for a pump shutdown with the loss of cooling (high engine temperature would be an easy one to use). No low tank level shutdown of any type either.

Once the engines fail, anyone in the pump room is at risk.

5 feet is probably safe but has anyone tested a system down to the Anti Vortex plate level and is 4 feet more likely a safe value?

I don't currently know if the procedure will allow manual shutdown of the pumps prior to destruction.

The procedure does have an increasing level of reporting that we are getting critical as the tank level gets lower as well as instructions to evacuate at 5 feet currently if not ordered to shut them down.

Tank makeup water system does not have the refill capacity to maintain or even significantly assist the situation in a multi zone fire (manual bypass is part of the procedure to help as much as possible and becomes irrelevant by the time 3 pumps are running at 100%)

 

[LittleInch]

Wow, you really do seem to have an interesting system, what with this and the non recircualting pump tests.

You don't really ask any questions, but I'll see what I can do.

1) The connection inside the tank - can it be modified? 3 feet seems like quite a large height and if you can turn the nozzle down or otherwise make the vortex plate better then you stand a chance of getting more supply. The only way to see if it works is to go down to your min level at a steady flow and see if the pump starts to lose pressure then go a bit lower until the inlet starts vortexing. If you can see into the tank at the time then all the better.

2) With respect to the remainder of the points, fire systems are often a bit brutal in design as they tend to dislike electrical control systems and things that can go wrong / turn it off due to false signal. The key part of any fire system design is to determine what it is trying to protect and for how long. Remote locations either go for a minimum get everyone safe then let it burn down approach or provide sufficient continuous water flow to keep a fire at bay. Locations where you expect extra help (the local FD) and other sources of water are available (the sea, river, lake, local water supply, fire tenders etc) then the installed fire system normally has a limited requirement (1 hour is typical). Your system seems to be somewhere in between??

3) Sounds like you need to beef up the re-fill capacity or the tank capacity to me. Has the plant got more zones since the fire water system was built?? or was it just inadequate from the start?



My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[LittleInch]

Also have a look at this site and follow some of the links

http://www.pumpfundamentals.com/help11.html

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Smokerr]

I think I obscured the question, it is what is the lowest level we can run before we loose suction in the tank?

I am having to review my data as I think the intake is 16 inches and the Anti Vortex plate is 4 feet (i.e. what happens when the water level drops below the plate. I need to translate my picture from the Pumphouse to my data sheet correctly.

I am following up on the link and it looks like it addresses my question, thank you, could find nothing.

In regards to the rest,

The suction is take from under the tank via a vertical penetration through the floor. Considering the freezing conditions it was the only way to do it rather than the preferred side presentation with the intake pointing down (which I would have though was the worst of the two so a new to me concept).

Ditto refill, design is supposed to be "adequate" but all too many stories of when systems were put to the test they did not work as designer thought they would.

Its an understandably tough one as not too many Hangars catch on fire let alone one that is close to your specific size and space configuration).

So the nominal 750,000 gallons is was not planned to be helped by refill, thats the FMX department looking at anything we can do to help out a given situation.

The same for pump shutdown, we loose the hangar and the pumps and then have to get an all new pumps as they no longer make the engine we have (it would be an ugly opportunity to get radiator fire pumps though!)

More likely we have a failure in the cooling system and loose an engine (and possibly injure or kill a mechanic) as there are multiple failure possible with that setup (we had a piece of wood from the construction show up in one pump which fortunately just shredded and plugged the strainer slowly and the heat build up was slow enough to be caught and FP shutdown before damage.)

I do understand about the shutdown issue, though in this case if you went with overheat shutdown thats on each individual engine via a mechanical temperature switch. Ironic they allow over-speed shutdown which is fully electronic triggered but not over temp.

And just for added fun, they decided the day tanks were too low and jacked them up in the air (top at 12 ft). Engine mfg was not present at the meeting to tell them it was not recommended. We have had the back checks fail 3 times putting diesel in the crankcase. All because of a code requirement about the fuel levels being above engine levels. I am not sure if other injection systems would do that (no engine mfg recommends it), the Detroit Diesel 2 cycles sure do not tolerate it with out solenoids and back checks (adding failure points to solve a non existent problem or something that the engine needs.

 

[Artisi]

First up,fire pumps shouldn't shut down under any circumstances, therefore if you reach minimum water level in your tank and the fire hasn't been extinguished then you have a big problem, overheating of the engine doesn't really rate as a major problem at this point.
From memory and it's many many years since I was involved with fire pumps the FM code doesn't allow for or specify any "shut-down" from high temp. low oil pressure etc.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[JJPellin]

I would not expect an engine driven fire water pump to fly apart on loss of cooling. The engine may eventually sieze up. But I see no reason why the pump or coupler would fly apart. The fire is likely the greatest risk. Run the engine to destruction and then buy a new engine.

Johnny Pellin

 

[Artisi]

That is the idea, run the pump until destruction if necessary, putting the fire out is number 1 priority. However you should ensure that you get the maximum amount of water from your tank before the pump goes off prime, so attention to the inlet is critical.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

smokerr, you need to send us a sketch of the intake as I can't figure it out. However even if you have a new design, can you actually change it? If you could fit a tee with two elbows pointing down and leave a gap equal to the square area of the pipe, but in circumference, then that is probably the best you can do, but you then risk sucking up any sh1t that has accumualted at the bottom of the tank. Solve one issue and create a few more...

How low can you go?? Depends on size, orientation, flow rate, but 1 foot would look to be the absolute lowest at full flow before you started sucking in serious quantities of air.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[bimr]

You can calculate the minimum submergence with the Hydraulic Institute's method:

http://www.pumps.org/content_detail.aspx?id=3524

 

[Artisi]

with an upward pointing inlet pipe, the minimum submergence before entraining air is anyone's guess.
an option may be to fit a "T" at the top of the upward pointing inlet pipe with 90degree bends fitted and facing downwards from the "T" to within a 1ft or so of the tank bottom - you may well need flow straighteners in the inlet of the 90 bends to cancel any likely pre-rotation leading to vortex formation- this might give you a minimum of a couple of feet of submergence - of course this depends on pipe sizes and flow rates - which so far have no been advised.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[ccfowler]

If I am understanding your situation correctly, the most important consideration is the avoidance of vortex formation at all cost. An important element in this is avoiding anything that causes a rapid acceleration of the flow anywhere in the tank system. Once formed, a vortex can persist for a very long time within the water in the tank. It can circulate away from the suction line an then return to enhance the suction condition problems some time later. When possible, avoid drawing water from the tank in surges, rather try to establish a reasonably stable rate of flow and hold that as reasonably constant as practical.

Pay close attention to the geometry of the intake so that water is drawn from the relatively motionless bulk of the tank through an intake configured to very gently accelerate the flow into the suction line. Usually this is most readily attained by a radial horizontal inflow that is gently redirected into a vertical suction line. Done properly, the water level can be drawn down to within a few inches of the bottom of the tank without inducing any nasty vortex formation troubles. Of course, when you run out of water, that situation will become very obvious very quickly, but by careful design of the intake geometry, a much greater portion of the tank volume can be used effectively.

In my experience, most of the intake designs in tanks are poorly configured with respect to permitting reliable withdrawal to very low tank levels. Any baffles in the tank could be helpful or harmful depending on their configuration and effects on flow within the tank. Protection from turbulence and vortex formation from the tank filling system must be given all due attention, too.

Of course, this is presuming that you are dealing with a reasonably clean tank with clean water. Commonly, intake design must consider problems introduced by trash or debris such as may be found at a pumping station drawing water from a river, lake, or sewer system.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

 

[Smokerr]

My apologies for not getting back to this, sudden heat wave (by our standards) and scramble on A/C as well other projects.

I appreciate the thoughts on this.

I am trying to understand any dynamics of the tank performance at low levels.

There is not intent to try to correct anything. .

I may have mixed things up as to what the question is. What I need to know is what happens as the water level gets down to and then below the Anti Vortex plate (if that is possible or if its just an unknown).

More specifically at what point with our setup at what point do we loose suction?. Its a vertical penetration through the bottom of the tank as follows.

Intake: 26 inches above the bottom of the tank. It has a bell on the end of the intake that goes from 30 inches to 39”. The bell occurs inside the tank in a length of 20 inches)
Anti Vortex Plate: 46 inches above the bottom of the tank. Picture a steel table with 4 legs that sits over the top of the intake. Not attached to the inlet.

20 inches of space between the Anti Vortex plate and the intake mouth.


I tried to follow the links to Gould info and they are broken. Have not been able to follow Goulds lit links through and find the right information. Would you have the direct link?

I don't know how to post a sketch and have limited resources in that regard as far as computer tools go. Give me a number and I can fax one. The link describes our setup as the second option in 2.12.1.6 . i.e bottom entry with intake vertical in the bottom of a tank)
(

http://www.fmglobal.com/assets/pdf/fmapprovals/4020.pdf‎)

Standard cylinder water tank (not on its side). Suction line goes straight down 12 feet from the tank and maintains that over to the pump house as we can get –20F here. Freezing is a given and that avoids the issue (possibly as some loss of tank volume).

Not first choice for Anti Vortex but listed as allowed and for our climate it makes sense (no heating systems to fail as freeze protected by depth of burial).

note: Overflow is a vertical tube in that tank that also goes through the bottom of the tank as well and then into the storm drain system for the same freezing reason. Otherwise we would have to protect piping from freezing and having everything go through the bottom of the tank avoids that (I think it’s the right approach as there is zero maint and monitoring with it and we have enough problems already)

We do keep the tank topped up as much as possible keeping in mind its an auto refill and the control valve drifts a bit. The nominal tank sizing gives us ½ ft lee way below the overflow.

And I am going off topic in responding to the following. Call it a severe disagreement between the maintainers and the committees that dictate the details. They don’t have to maintain it let alone pay for the consequences of what they cause.

I would not expect an engine driven fire water pump to fly apart on loss of cooling. The engine may eventually seize up. But I see no reason why the pump or coupler would fly apart. The fire is likely the greatest risk. Run the engine to destruction and then buy a new engine.

As mechanics, our entire existence is based on taking care of equipment. This is a case where we are told to let equipment self destruct let alone knowing that its for no purpose. A detached committed can put out an edict, but when your life is spent dedicated to doing just the opposite its not only difficult, it makes absolutely no sense with multiple pumps that have redundancy.

And unless trained and then drilled in, most people are going to try to save a piece of equipment as they are not thinking about their personal risk. That’s part of what I am working on is to get that established.

As a former care taker of the facility as well as the engineer now, I am obligated to ask the questions:
1 do we want to destroy an engine accidentally if there is a failure on the cooling system?
2. Can we prevent it? (we could destroy an engine testing for nor purpose as well)
3. Is there provision in the code to accommodate a situation where engine specific shutdown can be done when you have redundant pumps?
4. If engines are going to self destruct, what do I need to do to ensure the safety of the mechanics (or others) who have responded to the Pumphouse?

Most people have seen a vehicle with a blown radiator stranded by the side of the road. What happens is on loss of cooling (typical a broken fan belt ) the cooling system blows through the radiator cap causing that big cloud of steam. Engine is idled immediatly as the driver pulls off.

Engine is isolated form the occupants and the worst you get is nasty odor of anti freeze.

I know what happens when oil pressure fails, both at idle and under speed. Idle the engine stops, at speed the bearings spin and then things come out the side of the block.

Most people do not know most heavy duty diesels instead of a bore have sleeve inserted in the block sealed at the top and bottom with an O ring. That sleeve is relatively thin and its surrounded by the coolant. The coolant is going to overheat extremely quickly. Typically you have 30 seconds after a coolant failure before the engine destructs if under load.

Factors:
1. In this case, the load disappears (no water in the pump casing)
2. Various Overspeed, undershoot and recovery as the governor attempts to maintain 2100 rpm, so the engine is still at high speed and governor will keep adding fuel to maintain that speed.
3. Massive amounts of torque available as well as 500 hp or better.
4. Inertia from the rotating mass of the impeller/shaft assembly
5. Diesel engine with loss of coolant at speed and under load will seize in 30 seconds.
6. The heat exchanger immediately has no cooling. The closest vehicle analog is a fan (serpentine) belt failure)
7. Spewing of exploding coolant around can get into two intakes (its own intake and adjacent engine) and potentially hydro cylinders.


Initially there will be a steam explosion. Scalding and third degree burns if the fluid hits you. That is an incredibly serious amount of energy being released.

Actually engine failure I do not know. It may seize or it may mechanically self destruct.

If it can rip the rings apart and not seize up, it will continue to run until things come out the side of the block.

Also in play is the coupler and the dynamics of the engine and the rotating mass of the impeller. The mass will assist the engine in maintaining rotation if it does start to seize and the whole thing may have a coupling failure of its own.

Regardless, you do not want to be anywhere close to the engines when the failure occurs.

NFPA can be very intellectually detached as they are not in the room, but there are consequence beyond just the engine stopping. One of the ones I was called in on afterwards also caught fire. That and the others throw stuff around the room or engine compartment (connecting rod failure).

Supposedly the main goal is life safety followed by property protection. My take is that life safety also includes the pump house and safety of the mechanics who are tasked with responding. As the main looses suction its going to be a string of six steam explosions followed by a destroyed engine failure.

Even the redundant pumps will go as loss of main pressure then triggers them to run despite the fact there is no water to fight a fire with anymore. As they would have less load (no water to start with) it would probably take them longer.

My experience with NFPA is that they have become fixated on single fire pump situations and applied that to multiple fire pump situation and a lack of understanding in general of engine driven fire pumps (they are a small minority of the installed systems).


In this case with redundant pumps and their own sensing, there would be no reason not to shut an engine down. There would be no common point of failure as each engine has its own temperature (and oil pressure sensing). Overheat is going to destroy the engine and potentially impact the entire system at no benefit. A dead or even damaged engine is not going to be able fight a fire.

And the fact that they have over-speed protection shutdown (not just blow-off valve) shows that there is some recognition that if it endangers the system you not only can but are actually mandated to shut it down. They recognize the hydraulic implications but not the engine related ones.


These should be treated as systems, not individual components that happen to be put together and a thorough understanding of all the components is required, an engine is not an electric motor.

In this case of multiple pumps with redundancy, an individual pump shutdown on a fault would be safer both to people as well as the best possible insurance that the system as a whole would continue to provide the structural protection its intended to.

While much rarer, oil pressure loss is guaranteed to be worse. In that case the bearings spin, the rods break off the crank and or the pistons wrist pins and those parts will go out the side of the engine.

It really exists a basic lack of understanding of engines when assembled with fire pumps (some very good sprinkler system engineers I have worked with did not have a clue on how engines work and the implications). Great answers on the hydraulic and huh on the engine.