Hot Water Circulation Pump.

[S.G]

Dear Team,
Please find the attached datasheet for a hot water circulation pump.
We have experienced failures and loss of metal in the impeller, as well as in the first 5 inches of the discharge piping.
The operating conditions are as follows:
• Suction pressure: 1.2 barg
• Discharge pressure: 4.9 barg
• Temperature of fluid: 100°C
• Flow rate: 180 m³/hr to 200 m³/hr
Could you please confirm whether we are operating the pump outside its recommended operating envelope?

 

[georgeverghese]

Am not a water treatment person, but from what I gather in a few minutes from the net, max permissible dissolved O2 in closed loop hot water systems in contact with carbon steel is 100ppbw to reduce corrosion. To be safe, say operate at max 70ppbw to allow for analysis errors. Use metabisulfite if bisulfite is not satisfactory, or preferably use hydrazine. Maybe also ask some one to check N2 blanketing gas quality at expansion drum. TDS seems very high too for what should be demin water.
Are you using some kind of pressurised shaft seal at these pumps ? Is this pressurised flush liquid adequately de mineralised/ de oxygenated ?

 

[pierreick]

George,
I believe the head loss on the suction line is much higher than 0.4 bar given by the gauges, my calculation with a 10 m length pipe +tee and Check valve, diameter 80 mm is about 1.5 bars, meaning that NPSH a much lower, explaining cavitation.
Increasing the DN from 80 mm to 100 mm you can recover 1 bar!
Note: Your comments about the quality of the water are right, where did you get the info about Demin or Deionized water, well known to be corrosive? In this case the TDS value is abnormal, cross contamination for sure.
BTW Carbon steel is not suitable for corrosive water.
Pierre

 

[S.G]

Dear George and Pierre,

Attached is a snapshot from the 3D model. I will also send the piping GA shortly.

 

[pierreick]

Hi,
Attached a link to the details of the pump designed for hot water, supplied by KSB :

https://image.indotrading.com/co48959/pdf/p452564/6aeacffb-e9bc-43ae-8c5c-00de14c47b99.pdf

Pierre

 

[LittleInch]

Pierre,

The 3"/80mm line is from the make up tank to the pump circuit, NOT the main header etc which is 12" and 14" in the headers.

It is clear now though that the flow figure are independent of what the individual pumps are doing as the flow figures listed are from each hot water system which uses the common line from the pumps.

My question now is whether there is in fact some temperature or flow control or cross connection between the inlet line from the pumps and the common line from the heaters to the HX. Or a valve is passing allowing more flow from the pumps than is being recorded by the heaters??

As we appear to be getting only one piece of information at time this may take a few days before we get the WHOLE PICTURE....

If we say the max inlet temp is 105C ( this figure has been wondering around all over the place)we're at 1.2 bara / 14m given the lower SG.

The inlet pressure at the pump is now listed in worst case at 0.8 barg, say 22m. Minus 14m is 8m. NPSHR is 7m, but cavitation is higher so I still think the primary cause is cavitation. Plus that DP tells me there is more flow going somewhere than just your heaters so your NPSH/cavitation limit is going up.


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

 

[pierreick]

Hi LI,
My calculation is related to NPSH available where the suction line is 3".
In other words, the NPSH available is less than the NPSH required.
All the calculation are showing cavitation, design issues (velocity) and pump selection.
I still don't understand why KSB was offering pumps running at 3555 rpm, I'm more familiar with 1450 rpm for this application.
@ SG, I cannot offer more.
Good Luck
Pierre

 

[LittleInch]

The main suction header is 14" feeding four pumps.

Each pump suction line is initially 8" then reducing to 4", with a 3" pump outlet going back up to 8" before hitting another 14" header.

There is no flowing 3" pipe.

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

 

[S.G]

Dear All,

Please find the complete set of P&ID documents attached. We are currently awaiting feedback from the chemical supplier regarding the recent failures. Notably, these issues arose following the commencement of a chemical injection in August 2023.

It is important to highlight that this system has been operational since 2014 without any issues. However, a trend analysis indicates that the system has been running at higher suction pressures (1.5-1.8 barg) and higher discharge pressures (5.5-7 barg), with a minimum Delta P of 5.5-5.8 barg.

To address this, we are planning to adjust the PCV set-point to 2.5 barg.

Best regards,

 

[pierreick]

Li,
the discharge pipe from the pot is 3".
Anyway,I'm out.
Pierre

 

[S.G]

Hi Pierre,

Thank you for all your help and input. It is greatly beneficial to have such discussions.

 

[3DDave]

If the metal is affected by the chemicals I would expect a high amount of dissolved metal.

If it is cavitation I would expect an amount of metal particles, somewhere.

 

[LittleInch]

Sg - the file simply say "P" so won't download.

The other option is that the non working pump non return valve is passing and flow is recycling through it. Try manually isolating the non working pump and see what happens....

"However, a trend analysis indicates that the system has been running at higher suction pressures (1.5-1.8 barg) and higher discharge pressures (5.5-7 barg), with a minimum Delta P of 5.5-5.8 barg." I told you - SOMETHING else would be different. Somewhere the flow is bypassing between the discharge line and the inlet line causing the pumps to over flow, increased flow through the pump and increased NPSH / cavitation limit. It's either the pump or some other cross connection is open when it should be closed or is passing. Do a very thorough check and close all valves which are marked NC and try closing manual valves where there are NRVs to see what happens.

Increasing the inlet set pressure will help, but your pumps are running off their curve with a 4 bar DP so they will continue to fail seals, bearings and possibly cavitation damage.

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

 

[3DDave]

It's the "&" in the file name. You can copy/paste the link and replace the "&" with the unicode number.

I see it's removed from the latest attachment.

 

[LittleInch]

Well its a fairly simple system so the key issue for me would seem to be the standby pump NRV passing.

There is potential for some 1" valves to cross connect the two headers, but that shouldn't have the sort of impact we're seeing.

Its either that or all the existing pumps have been eaten away as well so they are not operating the same as they were.

Also have a check on power consumption. Not sure what they should be doing, but about 41kW looks about right. If they are drawing more current then they are working harder than they should be.

Also see if you can get a portable ultrasonic flow meter and see what each pump is doing.

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

 

[georgeverghese]

@pierre,
Hot oxygenated DM and DI water is quite corrosive, DI more than DM. This is why oxygen scavenger injection is critical in DM / DI hot water closed loop heating systems to remove dissolved O2. Pls see internet literature.
The main flow of HW is through the 14inch return header. This 3inch line from the expansion drum exit is only for pressure equalisation at the return header.
@SG
a) Make up water tank (which feeds into your expansion drum) has an open vent to atm. This is rather careless design - it should be N2 blanketed. And it looks like the feed to the makeup water tank is from the fire water tank ! This surely cannot be biocide doped demin water, and this firewater must be crawling with algae and bugs. This explains why you have 800ppbw dissolved O2 in the expansion drum.

b)The suction to each of these HW pumps goes down to as small as 4inch for 180m3/hr - I cannot make out the length of this 4inch section downstream of the suction strainer from the 3D model picture. Line velocity here is 6m/sec, and dp = 3.3bar/100m. Have you done a detailed dp check ?

c)Just for the sake of completeness, pls also post the drawing 10-PR-PID-0011, which shows the crude/ HW exchangers 20-HE-02A/B. Presumably, this dwg would show the overall min flow recycle loop for the HW circuit also ( I dont see this FIC recycle to HW return header in any other dwg). Some designers show this a constant dPIC control loop.

No explanation given so far why suction pressure was dropped from 6barg to current 1.2-1.5barg, but I now suspect this is because your plant integrity mechanical / corrosion engineers have discovered there is significant metal loss in the HW heating pass coils in the hot water generator, and possibly in the CS tubes of the crude / HW HX also.

 

[pierreick]

Hi George,LI ,
Thanks for the feedback, I should get a new pair of glasses.
Pierre

 

[S.G]

Dear George,

Apologies for the delayed reply. Please see the attached P&ID as requested. Also, I will check the length of the 4" section with the operations team. I have also confirmed with our operations that the internals of the check valve on the hot water expansion tank were removed prior to the commissioning of the system in 2014.

 

[S.G]

Dear LittleInch,
The power consumption stands at 36 kW. Furthermore, please note that the discharge and suction manual valves for the non-working pumps are always kept closed, in accordance with our operating procedures.

 

[S.G]

Hello George,

The length of the 4" section is 800 mm.

 

[georgeverghese]

Head loss here, inclusive of the upstream reducer is 1m, so total head loss = 2m. We still have more than adequate margin to NPSHr.
At 36kW, this sits well on the power vs flow curve for this pump.

So you have this 3way control valve on the hot water feed to the HX which is doubling up as min flow control valve for the pumps. Duty here is 10MW each, so both units running to get to 20MW design case duty? What is design case duty for each hot water fired heater - 10MW also?
You've declared pH adjustment and dissolved O2 scavenger chemical injection since Aug 2023 so far. What about corrosion inhibitor or antiscale inhibitor or biocide?
The unit was shutdown for 6months. Was the unit thoroughly dried and N2 blanketed during this time, including all of the HW side?