Engineers needed. Question about fouling factors in heat exchangers

[king of nowhere]

I admit, it doesn't have much to do with kerbal space program, or with space in general. But this forum is the best collection of nerds and stem-oriented people I know, and it was exceedingly useful in providing answers when I had an issue with cryogenic cooling. Besides, I'm doing this for a group of kids. Somebody think of the kids!

I am a high school teacher of chemistry, preparing a lesson on heat exchangers. My university curriculum was more on organic chemistry, I'm not a huge expert on the topic, and the textbook is not helping.

So, fouling factor. Your heat exchanger is going to get dirty, so you have a fouling factor to deal with that. it's a simple addition to the heat exchange coefficient.

However, the problem I have is that I expect fouling to build up with time. I would expect a fouling factor to have a buildup per day, showing how your exchanger becomes less and less efficient with time, until eventually it's time to clean it up. You could also use one such formulation to determine heat exchanger size depending on how often you want to run cleanup. But no, fouling factor is a simple factor that doesn't account for time at all.

I've been looking in the internet, but that's the kind of very specific stuff that's hard to find. I found lots of textbooks, but none mention time. The best I got is this quote
 

Quote

 

Practical design of heat exchangers

Plate heat exchangers are often an economical solution if corrosion resistant materials are required. The deposit forming is usually considered by calculating an excess heat transfer surface. The size of the excess surface margin must be as high as the heat exchanger can be used a reasonable time.

TEMA shell & tube heat exchanger

For the calculation of the initial excess size of shell & tube heat exchanger is the guideline with the fouling resistance terms of the Tubular Exchanger Manufacturer’s Association (TEMA) useful.

Other heat exchanger constructions

Plate Heat exchanger Caused by a higher turbulence plate heat exchangers leans to a lower deposit forming as shell & tube heat exchangers.

An examination of the deposit forming in plate heat exchangers and shell & tube heat exchangers with produced cooling water has the result that a plate heat exchanger with a velocity of 0.45 m /s has only one third of the deposit forming of a shell & tube heat exchanger with a velocity of 1.34 m/s. An examination with black liquor (67% solids content) shows a deposit forming in a double tube heat exchanger during 2 hours.

The parallel running plate heat exchanger was free of deposits after 24 hours. Nevertheless the heat transfer coefficient of plate heat exchangers e.g. in the dairy industry went down around 30% during one day.

Caused by hygienic reasons and the resulting recurrent cleaning cycles is fouling in this kind of industry a secondary problem.

Typical heat transfer coefficients of plate heat exchangers are around 6000 W/m²K compared with this are the heat transfer coefficients of shell and tube heat exchangers around 1500 W/m²K. Caused by these different heat transfer coefficients it is not advisable to calculate plate heat exchangers with the fouling resistance of shell and tube heat exchangers. With an exemplary fouling resistance of 0,15m²K/kW is the surface margin of a shell and tube heat exchanger around 22%. For the same example the surface margin of a plate heat exchanger will be 90%.

There are different recommendations for the fouling resistance of plate heat exchangers. In general the choice of the fouling resistance should result an over sizing of the plate heat exchanger below 25%.

 

underlined by me the parts that are more useful.

this piece acknowledges that time is an important factor, and that the exchanger will become less and less efficient until it has to be cleaned up. However, it doesn't in any way help with determining how often you have to clean it up.

It's like the fouling factor assumes that you are going to clean the exchanger regularly, before the efficiency drops even lower than the factor would suggest. But it doesn't in any way help determining how often that cleanup must occur. Once per day? Once per week? Once per year? I would assume that in the dairy industry you will need daily cleanings for hygiene purposes alone, so it's pointless to design a heat exchanger to work more than 24 hours without maintenance. But in the oil industry, you'd rather keep the machinery running for as long as you can, having to stop a whole distillation column daily to clean it would be unacceptable, and therefore you'll want them to run longer; possibly as much as a year, so you can shut down the plant during summer for maintenance, send most of the workers in vacation, and have the maintenance team clean. But how that affects fouling factors? are the fouling factors for food liquids calculated over different times than those for oil?

or does fouling eventually stabilize to a certain level of inefficiency? is the R factor a kind of asymptote to which the fouling tends? but in that case, why do regular cleanups, if it's not going to get worse?

The way the textbook simply dismisses the problem by giving the formula with the fouling factors is extremely confusing.

For the sake of preparing a better lesson for my kids, I hope in this forum there's some engineer with specific knowledge in the field that can answer.

[Codraroll]

My best guess: it's a problem that needs to take into account a lot of factors for accurate calculations, and many of those factors can be hard to obtain good data about in operation. That means the calculation has to rely on assumptions piled on assumptions until you can't really create a reliable answer. So the whole complicated problem, including the maintenance frequency, is simplified to a single factor and counted as a constant inefficiency. The average of a sort of "sawtooth curve" as the heat exchanger is fouled over time and cleaned occasionally. The factor you end up with could be based on some old research projects or a collection of industry data. Such things are done all the time in standardization, and it can be quite a rabbit hole to dig up the original source of a number.

[king of nowhere]

13 minutes ago, Codraroll said:

My best guess: it's a problem that needs to take into account a lot of factors for accurate calculations, and many of those factors can be hard to obtain good data about in operation. That means the calculation has to rely on assumptions piled on assumptions until you can't really create a reliable answer. So the whole complicated problem, including the maintenance frequency, is simplified to a single factor and counted as a constant inefficiency. The average of a sort of "sawtooth curve" as the heat exchanger is fouled over time and cleaned occasionally. The factor you end up with could be based on some old research projects or a collection of industry data. Such things are done all the time in standardization, and it can be quite a rabbit hole to dig up the original source of a number.

yes, that's my best guess too, but it should still include the information "how often are you going to clean up the exchanger". without that information, a fouling factor is kinda useless. and that information is not included in any source i could find, and even the best case only mentions that time is a factor without being any more specific.

P.S. you don't want to take the average of a sawtooth curve. you want to take the worst performance possible. because your chemical plant will have to keep working up to the point when you clean up the thing

Edited August 25 by king of nowhere

[darthgently]

Fouling covers a huge amount possible reactions and conditions: variances in the fluid composition, chemistry within the fluid across the temperature ranges in the exchanger as well as chemistry between the fluid and the materials that make up the exchanger.  I wonder if it's just too broad of a thing to nail down easily.  More of a "do it when it needs it" thing.  As an example I know that some well pump systems have seasonal variations in fouling as ground water can vary quite a bit seasonally with regards to microscopic stuff that lives in the water, clays, turbidity, rainfall, etc.

Somewhat, sort of, related, on a stretch, no hijack intended, DM any responses I suppose:

Spoiler

I've always wanted to build a very simple but very efficient heat exchanger for solar water preheating via a winter friendly medium like alcohol or antifreeze.  The design I've settle on is simply a very long length of copper tubing inside a very long length of a bit bigger diameter tubing (PEX?) with the fluids running in opposite directions.  I figured 3D printed spacers that keep the inner centered and not touching the outer tube while still allowing good flow could be figured out and the tubing could be loosely coiled and surrounded by insulation in a box

 

[mikegarrison]

Jet engines deteriorate with use. The blades get dinged, they get dirty, the seal clearances wear, etc. This causes degradation that slowly gets worse over time, but often is accounted for by a single fixed number when you are talking about fleet averages. Why? Because for economic reasons, when it becomes bad enough on a given engine, that engine gets cleaned. And if cleaning isn't enough, it gets rebuilt. So the fleet average hovers between "new" and "so degraded it is worth being taken care of". Which means you can sort of treat it as a constant factor.

Maybe heat exchangers are the same.

[Nuke]

i know materials incompatibility is an issue with water cooling loops. never mix aluminum and copper. with hvac exchangers filters are usually used to keep the exchanger bits itself clean. filters detract from performance but its usually negligible to the cost of replacing or servicing a dirty exchanger. 

[king of nowhere]

5 minutes ago, darthgently said:

I've always wanted to build a very simple but very efficient heat exchanger for solar water preheating via a winter friendly medium like alcohol or antifreeze.  The design I've settle on is simply a very long length of copper tubing inside a very long length of a bit bigger diameter tubing (PEX?) with the fluids running in opposite directions.  I figured 3D printed spacers that keep the inner centered and not touching the outer tube while still allowing good flow could be figured out and the tubing could be loosely coiled and surrounded by insulation in a box

 

you know, that's exactly the kind of problem my students are expected to be able to solve at the end of the topic.

except it misses several important details. namely:

- how much water you want to be able to heat? for a shower, you need a flow of 0.17 liters per second, if you are happy with slowly filling a tank, it can be less.
- by how much you want to heat that water? do you expect your design to raise temperature by 10°C? 20°C? that, and flow, will determine the amount of heat exchanged. of course, the upper limitation to that is the heat that can be generated by your solar heater.
- I'd also need the diameter of the inner copper tube
give me those data, and I can calculate something.

 

but as a very rough start, a system like that should be able to exchange roughly 1 kW of heat per square meter of surface of the copper tube per degree of average difference of temperature between the flows. I'm using 3000 W/m²K as convective coefficient for water, as a table i found claims it's a typical value for water in forced circulation. I'm using the same value for alcohol, because i can't find an appropriate value but it can't be too distant from water. you can use that to calculate how long the tube needs to be depending on how much heat you want to exchange