Diesel Driven Fire Pump

[hgordon]

Hi all,

I am troubleshooting an old diesel driven fire pump (sprinklers) which is giving less pressure boost at 5600 l/min than it used to.
In 2009 it produced 5600 l/min at 680 kPa at 1900 rpm, but in 2018 it produces 5600 l/min at 540 kPa at 1900 rpm.
Talking to a diesel guy, he says maybe the diesel has lost the torque it used to have (installed since 1980 Ford Leyland).
Which sounded logical - but would that not mean in 2018 at 5600 l/min torque load the rpms would decrease below 1900rpm because the torque is not there from engine to keep the revs at constant 1900 rpm?
Instead the revs are constant but the boost pressue is not - sounds to me more like to pump impeller is the issue?
Because the pump is still getting the 1900 rpm but not putting out the boost (impeller is always same).
This sounds right?

Thanks,

 

[hydtools]

Something changed the system load to reduce its pressure demand.if the flow rate is still the same.

Ted

 

[bimr]

Not really. But a guess should not be made with only partial information.

Please post the manufacturer's certified pump test curve. It should look like this:

The pump should be operating on the performance curve. If the diesel was operating at the same RPM's and all hydraulic conditions were the same, you should be getting the same flow. The output pressure of a centrifugal pump varies with the square of the speed. Pressure can be controlled accurately by controlling the speed of the pump. By adjusting the speed as the flow changes, the discharge pressure of the fire pump can be held constant for any flow.

The hydraulic conditions need to be verified. Are the hydraulic conditions such as water supply pressure the same as when the original test was done?

How much usage does this fire pump set have? In the event that you do have a fire and this fire pump set breaks down, a reasonable person would consider operating a 35-year old pump suspect.

 

[Artisi]

Same flow at difference pressure at same the speed on the same system doesn't make any hydraulic sense. Have you checked the pressure guage?

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.)

 

[hgordon]

Hi Guys, thanks for reply,
We change the system load by opening closing the valve to get 5600 l/min.
Thing is, it should be able to put out 5600 l/min with a discharge pressure of atlesat 550 kPa as it was designed.
This is so the sprinklers are getting the pressure they need downstream.
The pump comes stright from a 1m high tank and there seems to be no restriction as suction pressure only comes down to -5kPa. Which is negligble.
The original pump curve shows the 550 Kpa at the flow could be reached with about 1870rpm.
There is another primary fire pump off the towns main so redundacy there.

I am quite confused how it was putting out 680 kPa in 2009 actually as the pump will need to be at much higher rpms to achieve that - maybe some of this data is bad..
But I think I can rule out the diesel engine as it still gives 1900 rpm when things were good and now when things are bad..

 

[Artisi]

how are you measuring flow / pressure -- is it accurate ?

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.)

 

[hgordon]

Most of the data I have for flow and pressure is from surveys done every 2 years by sprinkler certifiers they use calibrated P gauges out of the box and flow gear that should also be calibrated. I also used my own gauges and two flow meters to get some of my own data and found similar to theirs.
However I did notice the permanent pump discharge P gauge is bad and has been for years so is being calibrated.
But like I said for our tests we use our own gauges. I am not sure if the surveyors use tacho meters to check rpm though and I also have not.

I think I have convinced myself though.
The pump at dead head (no load) sits at 2000 rpm.
When the valve is opened up to allow flow the pump suddenly needs torque applied by the diesel to do this.
the torque at 2000 rpm is not enough so it slows down to 1900rpm where the torque is enough.
Same goes if the torque was lacking due to a worn engine - the rpm would need to lower even further for more torque to drive the load.
So I suppose firepumps always want to run in their power bands, max torque to max power.
However in this case the rpms always remains at 1900 rpm so I suppose torque is enough but the pump is not..

 

[Artisi]

how many hours has the engine run?
think you are getting too hung up on the engine, it runs at 1900rpm on load as it is probably set to run at this speed.

How do you know the head was 680kPa in 2009, certified gauge ?

is it currently operating as designed for the fire duty xxx gpm at yyy kPa - if yes what is the problem?.

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]

The way NFPA treat pumps is nothing short of scandalous IMHO. It requires the pump to run at full speed against a dead head for THIRTY minutes every week, or something like that. The energy of the engine needs to go somewhere, usually by the temperature of the water and consequential cavitation and damage to the pump impellor.

What has been the normal testing of this unit?

After 35 years of doing that I think if you pull the pump impellor, you will probably find out it isn't quite the same size as it was when it went in....

I tried once to make this test have at least have 15% of rated flow and got completely ignored because their design was "certified" by UL or similar and they wouldn't change a single thing.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[hgordon]

Here in New Zealand we have our own Sprinkler Standard - but if our standard is missing something we look into NFPA clauses.

To test our pumps every week we run them through a test return line back to tank or recirclation line with orifice plates sized to induce the pump duty. Orifice plates should be sized to induce 110% of the duty flow.
However I have seen some testers just run them against dead heads. I have also seen them run the pumps through a test return with no orifice or back pressure so the pump tries to run way off its curve. I'm quite new to this industry and it is quite rough n' ready.

So for this site, the tester has been running the pump at like 180 kPa (no back pressure) at 1700 rpm so it is way to the right of the pump curve. Must be putting out a lot of flow too.
So I will be installing an orifice plate to replicate the pump duty.

The pump itself looks to have been replaced in 1996. But that is still 22 years.

 

[Artisi]

"So for this site, the tester has been running the pump at like 180 kPa (no back pressure) at 1700 rpm so it is way to the right of the pump curve. Must be putting out a lot of flow too."


How about a surprise and show everyone the pump curve - for the new pump of course

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]

So 20+ years of running way to the right of the curve, possibly / probably cavitating for 30 minutes a week?

Like I said - pull the pump impellor and have a look at it. If it's all there I will be very surprised.

but you're right, the way the fire protection industry looks at and treats pumps is not conducive to long term good operation, but I guess they work on the principle that it only needs to work once. Trying to modify it though is like trying to re write the Dead sea scrolls...

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.