7 replies to this topic

[PaulC]

Hi all,

I am interested in investigating the fouling factor currently experienced by one of our PHEs. Currently we clean the exchanger periodically (Every 9 hours). The idea is to have our PLC log the right parameters to be able to calculate the fouling factor and then clean when it reaches a certain factor rather than periodically.
Currently we are logging the inlet temperature, outlet temperature, heating medium temperature (~constant across plate) and the mass flow.
With this I have put the following calculations together
Q = m Cp dT which can give the duty currently experienced by the PHE, this is then put together with UA = Q / LMTD to give UA.
The idea is to log UA which will give an indication of the fouling factor as the only change in resistance in U will be the fouling factor on the PHE (Area remains constant). Hopefully when we have logged the data for a while and gone through a longer trial runtime we can determine what a high and low UA is.

Does this seem a reasonable way to determine the fouling factor for this PHE?

[breizh]

Hi Paul ,
I don't think the technology you are using is the right one if you need to stop and clean your equipment every 9 hours . I'm curious to know how you are doing the cleaning , to me the only effective way for a PHE is to open it and clean every plate one by one !
Have you ever thought about spiral heat exchanger ?

My 2 cents.

Breizh

Edited by breizh, 31 October 2012 - 07:39 PM.

[PaulC]

HI Breizh,

We are in the food industry and this is used for heating a sugar-milk solution. The cleaning takes place through cleaning in place (CIP) (unit goes offline and is flushed through with acid to remove milk stones and then caustic to kill of microbes). They go through periodic opening and cleaning operations but on a day to day basis this is the main method. The main reason for the frequent cleaning is due to quality requirements, after looking into this unit however it appears that the process fluid can be kept far longer than expected and therfore we do not need to clean on a quality basis so frequently. This is why we are looking at the fouling factor to indicate when it may need a clean.

Do you think that the above method seems reasonable for measuring the fouling factor assuming that all other items in the U expression remain constant?

Thanks

[breizh]

Paul ,
yes you can but I 'm still thinking that your technology is not the best .

As you are working for the food industry you might be interested with this link:

http://www.nzifst.or...ology/about.htm


Good luck.
Breizh

Edited by breizh, 02 November 2012 - 06:15 AM.

[S.AHMAD]

PaulC
1. The approach of using UA is ok as long as the throughput remains constant over the period since U is also a function of flowrate (or Reynold's No.).
2. If the flowrate is changing, appropriate correction may need to be applied in order for the UA to be consistent basis and hence gives meaningful indication of fouling
3. Periodic cleaning of exchangers normally dictated by economics. For critical services, a spare exchanger possibly justifiable to minimize loss of production.

[PaulC]

Hi S.AHMAD,

The throughput changes quite often with fluctuations on an hour to hour basis. From the graphs obtained so far it does seem to change significantly in time with the throughput, would you have any suggestions as to how this could be accounted for? Could the fouling factor be standardised with the flowrate to give a more meaningful result?

Apologies for not stating earlier that we do have a PHE in paralell so that we are always running.

Many thanks for your help,

Paul

[S.AHMAD]

1. What I have done in the past is to calculate individual heat transfer coefficient to determine the clean overall heat transfer coefficient.U_c
2. The dirty overall coefficient is calculated from operating data U_d.
3. Then calculate the fouling factor R_d = 1/U_d - 1/U_c.
4. However, do energy balance first and you may need to reconciled the energy balance to account erroneous in flowrate and temperature data.

Edited by S.AHMAD, 08 November 2012 - 06:53 PM.

[PaulC]

Hi S.AHMAD,

Thanks again for the reply, that sounds like an interesting method. We have our control system set up to log UA and are collecting data now. I am away from work until the 19th so will have to try then.

The flowrate varies significantly from hour to hour but often runs at the same rate for 3-4 hours at a standard level (~12000kg/hr) and so we were thinking that we could simply compare the flows from when the PHE is first activated to the next point when it runs at the same flow. A fully automated method would be better however so I will look into it further.

Thanks for all your help, I shall return with results in 10 days or so.