Q & A > Question Details
One of our condensers --using cooling water as coolant media -- is located at elevated position. We can periodically isolate and dismantle this condenser, and upon inspection , the tube side (cooling water side) of this condenser always suffers from signifcant amount of fouling.
One of our colleagues suggests we install an "inline centrifugal pump " on the cooling water supply line into this particular exchanger in order to increase the amount of water flowing through condenser's tube hence minimizing the fouling rate.
I'm a bit doubtful about this suggestion, as this exchanger receives the cooling water supply from network header, thus the amount of water supplied to the inline pump will still be the same as the amount of water supplied directly to the exchanger without inline pump. An inline pump, in my opinion, will only increase the inlet pressure of cooling water into this particular exchanger. In my opinion, any attempt to increase the discharge valve opening of inline pump cavitate the pump if discharge flow is higher than suction flow received from network header.
I would like to hear the opinion from experts about the inline pump of cooling water network.

Thanks for all..
The suggestion from Mr. Banik sounds interesting, and I'm going to evaluate it.
Anyway, I'm still curious with the case of inline pump installed in the cooling water supply line of an elevated exchanger, whether it will be able to pull more water supply from network.
My premises are :
1. Let's imagine an elevated exchanger is normally supplied with cooling water flow of X m3/hr.
2. The original supply pipe runs on the same elevation with main header of H m , then turning up towards exchanger.
3. If I reconfigure the supply pipe to turning down of H m below main header, then turning up again H m before further going up to reach the exchanger, the pressure profile inside this reconfigured pipe at elevation of H m will still same with pressure profile of original pipe at elevation of H m.
4. Hence flow of water in supply pipe no. 2 and 3 will still same.
5. If I put a pump in lowest section of reconfigured supply pipe no. 3, then the amount of water flowing into pump suction will still same X m3/hr.
6. As centrifugal pump doesn't suck, but it only pushes, so the amount of water pumped will still same X m3/hr. The only different thing is water inlet pressure to exchanger increases hence water outlet pressure from exchanger also increases.
7. Thus operating the pump discharge above X m3/hr will cause transient inventory loss in the pump casing hence cavitation.
Do I miss something or make mistakes in my premises above ?
15/01/2014 A: Mike Watson, Tube Tech International Ltd, mike.watson@tubetech.com
Shell and tube condensers can be cleaned mechanically online without the need to shut down. The system will keep the tubes clean and free from fouling. No chemicals nor filtration is needed.
29/08/2012 A: Ralph Ragsdale, Ragsdale Refining Courses, ralph.ragsdale@att.net
With adequate available NPSH, the pump will indeed “suck”, and not cavitate. Visualize the balanced pressure system described by Eric, with each user receiving a certain amount of flow. Then, start the booster pump. The pressures will rebalance, with changed flow rates to the users. The starved user will experience a new, higher flow rate. By starting with a pressure survey and subsequent calculations, as Eric described, one can predict the new set of flow rates with the booster pump running.
23/08/2012 A: S Banik, Centre for High Technology, sbdr@rediffmail.com
To increase flow of cooling water in the condenser, you can recirculate water from outlet to the inlet using a pump with proper instrumentation. Temperature of the mixed water would be higher than inlet cooling water depending on the quantity of recirculation. However it would improve heat-tranfer coefficient which may compensate reduction in LMTD.
23/08/2012 A: Ralph Ragsdale, Ragsdale Refining Courses, ralph.ragsdale@att.net
Carefully make the calculation for available NPSH and it should work without cavitation. In other words, if the inline pump suction is at the elevation of the exchanger, there are no values in the calculation for the static head from a liquid level or suction line loss from a liquid level. The important values will be the suction static pressure (pressure gage psia) and the vapor pressure of the water at that temperature. When determining the suction static pressure, allow for the higher line loss due to the higher flow rate.
23/08/2012 A: Eric Vetters, ProCorr Consulting Services, ewvetters@yahoo.com
Very often water cooled exchangers at elevated locations get starved for flow while other exchangers "steal" extra flow. One option to ensure that you get the desired flow is to install a pump as has been suggested.
Cooling water systems do not work as designed because there are no control valves to make sure that the flow goes where it is supposed to. Instead they work by hydraulics and the flow distributes itself however the hydraulics dictate. The hydraulics of the system are too complex and subject to change over time to have any hope of designing the cooling water system make the design such that the hydraulics balance to get the right flows to all the exchangers.
As a result some exchangers end up starved for water while others are "water hogs". The problems are compounded by poor heat exchanger design practices that result in low tube velocities. Low velocity and high tube wall temperature promote fouling, which further reduces flow and promotes even more fouling.
If you put an inline pump into the system it can ensure that the desired water flow (possibly more) reaches your problem exchanger. It works by pulling water flow away from other exchangers in the network. Remember the flow is dictated by hydraulics, not design rate to a given exchanger.
Before you go to this option you should do a cooling water flow survey to understand just how the flow is distributing in your cooling water system. This is typically done using a doppler type flow meter that can be strapped onto the outside of the pipe. Your water treatment chemical vendor should be able to do this survey for you as a routine service. You can also verify flow rates to individual exchangers by doing heat and material balances around an exchanger using actual process side flows and temperatures to calculate duty and calculating water flow based on duty and water delta T.
You should also look at your heat exchanger design. That may be the cause or contribute to your fouling. Ideally you would like to have cooling water tube velocity >6 ft/s. If you put in a pump you have the flexibility to change the number of tube passes to increase the velocity. Doing so significantly increases the clean delta P across the exchanger. For instance doubling the number of tube passes in a fixed size exchanger increases the pressure drop by a factor of 8x.