Heat Exchanger Efficiency

[1athoti]

Hi, I would appreciate some input or direction on some Heat Exchanger (HE) variables. I realize there are a lot of variables in this system, but if we'd at least begin with a few.
I'm dealing with a pipe heat exchanger immersed in a tank (about 600Gal of water) with a circulating pump. The HE is placed at one end of the tank, with the pump inlet at the other end forcing the flow over the heat exchanger.

Given a certain length of pipe, would I want more passes (in other words the same continuous pipe with more bends) or more pipes with fewer or no bends (more or less like a cast iron radiator)?
I expect more passes would allow more heat to be realeased into the tank, therefore HE delta T greater? Fewer passes would maintain the HE delta T smaller? This is considering the same length and size of pipe.

Thank you for your help!

 

[LittleInch]

Depends. Many variables.

The flow rate and temp differential is key.

The min practical DT at the end of your tube is about 5 to 10C. Any lower and the heat transfer drops to near zero. Any higher and you're not giving the fluid enough residence time to transfer heat.

So if your DT is high at the end point, then a longer tube would increase heat transfer.

If it's low then more parallel tubes with the same mass flow per tube ( so double or triple total hot fluid flow). Or increase hot fluid flowrate and have a longer tube.

Any specifics we can work with here?

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

 

[1athoti]

Hi @LittleInch, thank you for the input!

I have a boiler of 1.25mil BTU, though it is used for multiple machines, not just the one I'm working on. We can assume 1 mil BTUs since they don't all run at the same time. The pump on the boiler is about 80GPM flow. The boiler should be able to maintain a Temp of about 190F. We're looking at raising the temp of 600Gal of water from room temp (say 68F) to 160F so 92F DT in 30mins if possible.

At this point I have the heat exchanger designed with 1.25 OD pipes stainless 0.49 wall thickness.
There are 8 pipes of 45inch length going one way from the input manifold and returning to the output manifold. This brings one length of pipe to 90inch. The total length of HE pipe would be 720inch.

I could place a baffle in the manifolds to have 4 pipes and 4 passes, rather than 8 pipes x 2 passes. Would this be better? Thanks!

 

[EdStainless]

I am not doing the math, but do the BTUs work out?
Is there forced circulation in the tank?
If not I don't think that you stand a chance.
What are the materials of construction?
I hope that you are using the word 'pope' as a colloquialism, this needs to be thin wall tubing.
If you are trying to reach 160F then the hot water leaving should be 170-180F.
Smaller diameter tube will give you higher velocity and better heat transfer, but also higher pressure drop.
What does the pump curve from the boiler look like?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Heaviside1925]

OP,
I am not sure what you mean by "better".

You have a stated design condition of

We're looking at raising the temp of 600Gal of water from room temp (say 68F) to 160F so 92F DT in 30mins

Run the numbers for

There are 8 pipes of 45inch length going one way from the input manifold and returning to the output manifold. This brings one length of pipe to 90inch. The total length of HE pipe would be 720inch.

and

4 pipes and 4 passes, rather than 8 pipes x 2 passes

And see how long it takes to reach your design condition of 160F. I am wondering if you are getting hung up on heat quantity Q vs heat rate Q flux. Consider this is rate dependent problem and not just quantity. Not sure if that helps but I am still wondering what you mean by "better".

 

[georgeverghese]

This is a rather primitive way of heating this water. You'd be much better off with a HX outside this tank, where you could be running this pumped water stream through this HX. That way, you'd have forced convection on both sides of the HX which will give you much better heat transfer, rather than this current arrangement where you've got much poorer natural convection heat transfer on one side. Off the shelf finned tube units from Brown Fintube would be even better, and these folks can help you with selection also .

 

[Snickster]

There are only a few variable to be concerned with. For heat transfer across tubes:

Q = U A dT

Increasing any of the terms on the right increases the heat transfer.

U overall heat transfer can be increased by increasing the velocity in the tubes, but then pressure drop gets greater. It can also be incrased by incresing the velocity on the outside shell side of the tubes. Pressure drop also gets greater for shell side fluid.
A can be increased by using more passes in the tubes, or increasing diameter for same lenght.
dT is the log mean temperature difference and can be increased by increasing the LMTD. For a heat exchanger this is determined by inlet and outlet temperature so the heating and heated fluid and the configuration of the passes (i.e, single pass shell/single pass tube, or single pass shell or double pass tube, etc.). There is a formula for computing the LMTD for any configuration of inlet and outlet fluid flow.

Play with these factors to obtain maximum heat transfer based on limitations of your system.

Also flow of heating liquid makes a difference since heat content, and therfore temperature drop, and therefore LMTD is based on the equation:

Q = m Cp dT

Where m is mass flow rate
Cp is specific heat
dT is temperature difference of heating fluid inlet minus outlet

So the more mass flow of you heating fluid the less temperature drop and the higher the LMTD.

 

[Snickster]

Here is a good reference for designing shell and tube heat exchangers which is basically what your configuration is.

 

[1athoti]

Thanks everyone!

@georgeverghese, Yes, an outside shell and tube exchanger would be much more efficient. The issue, which I should have mentioned, is the fluid is caustic water which adheres / deposits to the tubes very quickly so the HE must be cleaned much more often. We use an external 890k BTU shell and tube with steam for a machine and it works great for heating, save for the cleaning issue. Another issue is, it’s difficult to filter the caustic water to help keep the pump and HE clean. The filters clog up very quickly.

One of the customers wants an internal HE which they will clean less often / at the same time the entire machine is cleaned and will not have to deal with filtering. I realize it’s much less efficient, but we wish to accommodate their needs. The tank does have circulation as I said at the beginning, a pump with suction on the opposite side of the HE which blows the water over the HE pipes. This should help but I’m not sure how much flow to figure for. In terms of surface area, the 890k BTU has 23 sq. ft of area and what I have now is 36 sq. ft (including the manifolds).

I’m afraid the fintubes would get cacked with caustic much too quickly or be difficult to clean, but I’ll take a look at those nonetheless. Thanks for the suggestion.

 

[Heaviside1925]

OP,
Each post you make includes more information. I am now seeing 3 different possibilities for "better". You have now introduced a fouling factor. I am assuming the working fluid is water and the fluid you are heating is the caustic water?
So let me ask you this, from your original design condition:

We're looking at raising the temp of 600Gal of water from room temp (say 68F) to 160F so 92F DT in 30mins if possible.

What is most important to you?
- Getting the caustic water to 160F the quickest?
- Getting the caustic water to 160F the most efficiently?
- The longest system run time before fouling reduces the efficiency to a point the HX needs cleaned?

 

[1athoti]

Thank you for the questions and I apologize for not giving more info to begin with. Yes, we're heating the caustic 600Gal water tank with hot water 190F. In this case it is with immersed HE inside the tank with flow only by a circulating pump, about 40GPM. There is no head to speak of.

The optimal would be getting the caustic water heated the quickest, within half an hour hopefully not more than 45min. We do not have issues with the external shell and tube HE at this point heating wise. I to achieve similar results with the internal HE and water. At this point I was at least shooting for significantly more surface area, 36 sq ft vs 23. There is fouling, but it can be dealt with. They use some acidic solution for cleaning off the caustic deposits. Actually, the biggest drawback to the shell and tube heater is not the fouling, but the filtering of the caustic water that runs through. The water is used to clean bottles, but it is not the final cleansing solution. As such, it's not 'sanitizing' clean. It may have sufficient particles to very easily clog up the filters.

I did run through some calcs and will redo, but on one hand wasn't sure how to estimate the flow in the tank. Then also how to optimize all the variables, since as was mentioned, the pipe size, velocity, pressure are all related. How do I mathematically achieve an optimal solution? Which of the variables should I make constant first? I apologize if this is basic information for someone knowledgeable - I don't deal with hydraulics and thermodynamics every day Thanks again!

 

[LittleInch]

The numbers here are pretty marginal at the moment.

I've done this in metric, but on the energy front raising the water the temp you need in 30 mins with ZERO heat loss to the outside needs a steady flow of about 271kW. Your 1MMBtu/hr is 293kW.

So not a lot spare for other losses such as heat int he condensate water coming out plus heat loss to the outside or air.

So 30 mins is looking like a stretch to me.
The other issue is that as you heat the water up your DT across your HX goes down over time so you need to at least use the average Delta T in your energy calcs. So your temp rise will be an exponential curve and your heat power going in will reduce, so look at the heat flow when cold and when hot.

I'm no expert on steam heating, but hopefully you get all the steam converting to condensate, but that condensate temp leaving the HX will be at about 5 to 7 C higher than whatever your water temp is at the time. Where does the condensate go? Does the condensation energy form the majority of the heat loss from the steam?

I have no real idea yet if your HX can supply the requisite power needed or not. I'm struggling to see how that much power can be transferred with that little length of tubing, but maybe it can - steam is a bit of a weird thing to me.

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

 

[Heaviside1925]

OP,
I mentioned fouling factor not as something to deal with but to consider in your main question, which is basically big HX area vs smaller HX area. Consider Q = U A dT. Fouling factor effects U, so if you have a larger A, you can negate some of the effects from fouling better than if you have a smaller A.

How do I mathematically achieve an optimal solution?

You can't right now with the information provided. You will need to experimentally determine U unless you have enough information to calculate it

https://www.tlv.com/en-us/steam-info/steam-theory/steam-basics/overall-heat-transfer-coefficient.

Consider your rate is in BTU/hr, so if you want the same exchange in 30 minutes...

https://www.pdhonline.com/courses/m371/m371content.pdf

 

[georgeverghese]

Suspect this caustic solution is made up with "regular" water, which would have its typical Total Dissolved Salts - major component of which is dissolved Ca and Mg salts. On contact with caustic, you get calcium hydroxide and magnesium hydroxide. Ca(OH)2 is only partially soluble in hot water, while Mg(OH)2 is near insoluble. Both will precipitate out of solution and bung up filters, pump, pump seals, piping and the HX. Use good quality demineralised water as solvent and this problem may resolve, and then you can go back to the more efficient external HX setup.

 

[pierreick]

Hi,
Please provide us with substantial information about the product/process, as the term "caustic water" is ambiguous. Are the raw materials a caustic solution or pearls? The process is known to be exothermic. Could you specify your goals regarding concentration and temperature, as well as the intended application of this solution? Additionally, is this a continuous or batch process? It is well-known that caustic soda presents several safety hazards, including the potential for severe burns upon contact.

Pierre

 

[pierreick]

Hi,
To add to my previous reply, Info about Caustic soda .
Pierre

 

[1athoti]

Sure. I am not a chemical engineer, so I slapped it under "caustic" and "acidic" solutions... It is a milk bottle washing machine, but I expect the same / similar solution is used for other bottle washing, such as wine or beer bottles etc. I think it is a NaOH, sodium hydroxide solution and it has been used for a very long time, nothing new here. The machine is used daily (or 5 days a week) for not more than 8 hours a day - it is rather on the smaller size, about 2500 bottles / hour.

I'm afraid I've sent you on the wrong path. The concern our customer has and the reason for the internal HE rather than shell and tube is not because of cleaning off the backed caustic solution, but rather the 'milk particles' and other particles. This tank is the soak tank for the bottles. After sitting in the tank for a while the milk becomes this sort of plasticky, stringy stuff which tends to clog up things pretty fast. Hence the need to filter the caustic water. We can deal with the caustic deposits, but indeed if we had an internal heater, we wouldn’t have to worry about filtering the circulating pump.

They’re set on the internal HE so we’ll give it a try. I agree the 30min is not sufficient and they’re OK with up to 1h. My main question was on the 8 runs x 2 passes versus 4 runs x 4 passes and I’ll have to revisit the calcs. Unfortunately, I’m pulled away from this aspect of the machine at the moment but will update when I get back to it.

I very much appreciate all the input!

 

[LittleInch]

You might need to switch between the two, so 8 runs by two passes up to to a certain temperature then 4 runs x 4 passes when you need a greater area to compensate for the reduced delta T? Maybe?

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

 

[georgeverghese]

Ok, things get a little clearer now.
Of the 2 heat transfer resistances on this U tube heating bank, the outside heat transfer coeff will dominate, since it is a lot less than that on the inside of the tubes where you have condensing steam at ??degC. So it doesnt matter which way you configure these tubes, since OD surface area remains constant, and OD htc remains the same. Changing configuration will not change steam condensing htc.
How about a jacketed vessel where steam runs through the jacket? Then you wont have to circulate this gunky caustic solution?