PHE Hot and Cold film Heat Transfer Coefficents 4

[SamOntario]

Basic overall HTC equation 1/U= 1/Uh + t/K + 1/Uc

PHE operation data: Primary side in/out = 120/70 degC flow=50m3/h softened water at 12 BarG, Secondary side in/out = 60/85 degC flow=100m3/h softened water at 10 BarG.

PHE construction material is SS316 t=0.7mm thick, thermal conductivity is approximately K=0.67 W/m/K.

I am puzzled that PHE manufacturer sales engineers giving me U ranging from 4000 W/K/m2 to 9000 W/K/m2 with the above data. When I ask for Uc and Uh, they could not provide and making excuses such as proprietary data. Pls advise how I could determine Uh and Uc within reasonable accuracy?

Pls also advise how to conduct performance tests if the operational data are deviated from the above design data, thanks.

 

[zerok]

Samontario,

there is a free software for plate heat exchangers design.

check it out at:

http://www.sourceforge.net/projects/innohex

 

[SamOntario]

Thanks for the link.

I downloaded and installed the software. It will not calculate Uc and Uh.

 

[rmw]

The plates for PHE's within a given frame size and type (narrow gap vs wider gap plates) come in a large variety of thicknesses, corrugation patterns, and inlet/outlet (on the plate) guide chanelling among other things.

The manufacturers sizing programs are going to select the most competitive plate size/number/combination for the offering that will meet the service within the constraints you dictate. (Temp drop/rise, flow rate, pressure ratings, etc.)

A simple change in the slope of a chevron type corrugation pattern can change the heat transfer characteristic of a given plate with all other factors held constant. Sometimes plates within a plate pack will consist of differing plate corrugation styles and chevron slopes in order to balance the aggressiveness of the plate design with the design parameters.

PHE mfg'r's have different 'factors of safety' that they use to insure that their offering will, in fact meet your service, varying from a negative factor of safety ("well, we can always add plates later if the unit doesn't perform, because we figure that the customer has already build into the design his own fat that the Hx will never really see") to genuinely positive factors of safety ("we are never going to be found not to have performed.")

They are not going to release the type of information you are seeking because it would jeprodize their competition find out their competitive advantages and/or their propriatary sizing criteria. They don't know that you are not their competitors bro-in-law.

And, eventhough a few mfg'r's stamp plates for a lot of their competition, they can't know how the final assembler of the Hx will mix/match the plates to arrive at a final design.

While the "U" offered does seem wide, did you carefully compare all aspects of the PHE's that you can determine, such as plate thickness, plate count, frame size, # of passes, etc?

Give your vendors some slightly differing conditions from those shown above, and watch the plethora of types and frames that will come from their offerings.

Such is the nature of PHE's.

rmw

 

[zerok]

SamOntario,

I assume that Uc and Uf refer to the heat transfer coefficients for the cold and hot sides.

The software does most of the calculations you need.

After entering all required data, navigate to the summary view on the case inspector pane and check heat transfer coefficients in the streams properties table.

the plate thermal conductivity you provided (0.67 w/m.k) is

abnormally low. SS 316 thermal conductivity @ 100 °C is 16.2 w/(m.k).

 

[speco]

Dear SamOntario,

The people who manufacture plate heat exchangers often do funny things with their calculations. They should be able to give you the component heat transfer coefficients for both sides of the exchanger. If they can't do this, they are either incompetent or lying. Lying is not out of the question, especially in the HVAC arena, where everything is driven by price. By the way, if they get caught cheating, they always have the option of providing additional plates.

Testing for an off-design case is a bit dicey, but the manufacturer should be able to calculate the expected results, and you can compare with you measured flows and temperatures.

Everything rmw has stated is true. There is a lot of variability between plate configurations, and you could easily get answers which are all over the place. There is one general principal that you can count on - The higher the pressure drop, the higher the heat transfer coefficient.

Regards,

Speco (

http://www.stoneprocess.com)

 

[rmw]

I downloaded the program given by zerok but did not have time to play with it. (in a moment, the star.)

I did notice that it had user selectable fouling factor inputs. I did not get a chance to study the impact of that feature.

I remember being at a technical seminar put on by a major USA PHE mfg'r 15-20 years ago or so for engineers from all types of processes in the region when they were still trying to firmly establish their niche in the heat transfer world fighting their arch rival the S&T's. Most of the engineers in attendance had extensive experience in S&T Hx design.

They hammered the point that while the selection of fouling factors was beneficial in S&T design to increase the overall surface area to account for anticipated and unanticipated fouling, (considered within the constraints of other design considerations, of course) that it was detrimental for engineers to specify a fouling factor to be applied to PHE's.

As speco so correctly states, velocity (which is a function of pressure drop) is everything to a PHE, and the addition of additional surface to provide for future fouling induced fouling because it slowed the velocity down and reduced the "scrubbing" action of the fluid as it made the tortuous path changes within the corrugations.

It also reduced heat transfer for the same reasons because it changed the film coefficients as the velocity changed.

I don't know if this meant that they had already put a fouling factor in their propritary sizing programs, or just ignored it. I have seen some PHE Mfg'r's whose PHE's were designed to handle nasty fluids (sugar, ethanol, etc.) include excess surface to account for inevitable fouling, but it was not done in such a way so as to 'baloon' the plate pack, but to 'stretch' the plate pack. That is to say, the flow path surface area was not increased overall by a 'fouling factor' but was merely made longer by additional passes or plate length.

It is something that has to be carefully considered in PHE application. They are great in their niche, but easy to mis-apply if you don't understand how they work.

rmw

 

[PHEs]

Dear SamOntario


Some of your PHE data are wrong, the heat in is the same as the heat out.The 50 000 (120 – 70) should be the same as 100 000 (85 – 60), so you have heat balance! One of your temperatures or flow is wrong. Probably the flow data.
When you have the right temperatures and flow, may be you will find that the manufacturer calculation is right.
Uc and Uh is difficult for you to obtain, but the manufacturer should have the data from test.
The formulas Q = A x K x tm can be useful. Where Q is quantity of heat transferred (kcal/h)
“ A “ heat transfer area (m2)
“ K “ “ “ coefficient
“ tm “ log mean temperature difference .


With kind regards

 

[SamOntario]

Thanks for the reply.

I have assumed water specific heat capacity to be 4.18 through out. You are right if you were talking about differences in specific heat capacities of water at different temperatures. The differences in load is trivial, say less than 5% when compared with what the manufaturer's quoted performance data. The manufacturers (6 of them), have quoted U to be 4000 W/K/m2 to 9000 W/K/m2 with resulting PHE surface areas ranging more than +/- 50% with same plate thickness, same material, under same operating conditions.

I ask them to provide the sizing/selection program but they are all reluctant to provide their programs. How can I trust these manufacturers, and who is giving the true performance data?

 

[25362]

This link may be of help:

http://www.muel.com/products/heattransfer/plate/hvac.cfm

 

[rmw]

Sam, there are so many variables that I can think of among 6 potential suppliers that they are too long to list.

Make a spread sheet and eliminate all the things that are the same among your bids, and then give us some information on what are the glaring differences are between their proposals.

For my part, I am expecting to see the same or similar plate metallurgies, and I would hope that they all adhered to your specified flow rates and temperatures, but I expect to see widely varying pressure drops and/or frame sizes and characteristics.

Some may not have a specific frame size that falls right into what you need, so they are pushing a smaller Hx real hard, or a larger one not hard enough, so to speak.

We can't help you without no more information than you have given.

rmw

 

[SamOntario]

RMW,

The PHE data from one of the manufacturer are as follows:

PHE1
Hot Cold
Flow 27012 54317 kg/h
Pressure drop 0.2 0.61 Bar
Specific Heat Capacity 4.212 4.189 kJ/kg/c
Thermal conductivity 0.678 0.664
Load 1580kW
HTC 9243.3W/C/m2

Plate 0.5mm SS316
# of Plates 51
Total active area 8.43m2

PHE4
Hot Cold
Flow 75222 151263 kg/h
Pressure drop 0.2 0.64 Bar
Specific Heat Capacity 4.212 4.189 kJ/kg/c
Thermal conductivity 0.678 0.664
Load 4400kW
HTC 6592.7W/C/m2

Plate 0.5mm SS316
# of Plates 62
Total active area 33.43m2

These 2 PHEs from the same manufacturer have identical conditions of operation, yet the HTC or U varies from 9243W/C/m2 to 6592W/C/m2. Its quite challenging to make sense out of it.

I would appreciate any comments from the above data.

 

[zerok]

SamOnTario,

Could you provide plate details for PHE1 and PHE4:

- Chevron angle.

- plates gap

- plate width and/or length.

- connections ID.

I'll use the software ,I provided the link to, to give you some performance figures.

zerok

 

[rmw]

Sam,

Sorry to be so long responding. I am travelling, and haven't gotten on line much lately.

But, now as to what you have presented. At first glance (and I reserve the right to revise and extend my comments) I see that your flows in the two comparisons are roughly double relative to hotside/coldside in the case of PHE 1 and PHE 4. The ratio of cold to hot is roughly 2:1 in each case. But note that the ratio between heaters, while the hotside of #4 is only double the hotside of #1, the coldside of #4 is nearly triple that of #1. The plate count is in the neighborhood of being the same or very similar (as opposed to a 2:1 or 3:1 ratio).

However, note that while the ratio of the duties is roughly 3:1, the surface area ratio between the two units is roughly 4:1 for what I consider to be a similar plate count, so it appears to me that there must be quite a difference in the relative size of the units, especially width, which changes the velocity through that similar number of plates and which tells me that you are not comparing apples to apples, so to speak.

Had you had for your flow conditions of roughly twice the flow on both sides for one heater over the other, and had a similar (frame) sized unit with twice the plates or so, then there would be cause for concern, but in your case, with 3 times the duty and almost 4 times the square meters of surface area for three times the duty, you can't make the comparison you are trying to make.

There are other things that govern PHE sizing, such as connection sizes, pressure drop limitations, etc., that sometimes drive the unit selection instead of purely plate surface area.

I don't think you have anyting to be concerned about. You should be evaluating on other factors such as pressure drop, $/square meter, $/unit weight, and so on rather than concerning yourself with the HTC.

Zerok, I ran a liquid/liquid case on the software you recommended based on a known exchanger, and the results were impressive. I also ran a case where there was a phase change on one side, and it won't do that.

Sam, They should have given you enough information, except for the angle of the chevrons to be able to let Zerok run a case for you.

Zerok, with all but the chevron angle from Sam, you should be able to back into the chevron angle with all the other known information.

Hope this helps Sam. I may see more as I look over this more later.

rmw

 

[SamOntario]

RMW,

The manufacturers will not provide this chevron angle.

You are correct that the above example was comparing apple to apple since the loads are different, the frame sizes are different. In the above example, it does not make sense to have same design conditions, almost same pressure drops, same plate materials and thickness, and yet, the same manufacturer for PHE1 and PHE4 has dramatic differences in HTC. How can I explain this to someone who will be operating them. There are 8 sizes of PHEs of various quatities to be purchased. The HTCs for them are all different from a particular manufacturer.

Sorry that I had not yet provided 2 different manufacturers of the same loading PHE that you have needed me to provide. I was really puzzled by the HTC provided by that manufacturer. I will provide them at a later stage. Can someone provide explanation on the differences in HTC in the above example?

 

[zerok]

Hi SamOntario,

I think the manufacturer should be able to provide you with the plate dimensions. I will back calculate the chevron angle if needed (normally between 25 and 65°).

Hope this helps.

regards

zerok

 

[rmw]

Maybe I didn't emphasize it in my previous answer.

It is obvious, now verified by you, that the frame/plate sizes are different. The fluid velocities (both hot and cold) in the two exchangers is different, hence the different HTC.

rmw

 

[SamOntario]

zerok,

Thanks for trying to help. In response to your previous thread, the followings are data provided by the plate manufacturer:

1. Chevron angle: Manufacturers did not provide, pls assume the best and worst scenario.

2. Plates gap: Manufacturers did not provide, pls assume a common gap.

3. Plate widths for PHE1: 353mm PHE4: 620
Plate Heights for PHE1: 1107mm PHE4: 1765mm

4. Connection IDs for PHE1: 50mm NB for all 4 connections
for PHE4: 100mm NB for all 4 connections


RMW,

Thanks again for your inputs.

 

[SamOntario]

zerok,

Sorry that data in 3. are all PHE overall dimensions which will be probably irrelevant to the plate dimensions you required for your using the software. Again, the manufacturers did not provide all plate dimensions, except the plate thickness.

The cold side has a high HTC and the hot side has a low HTC. When combining the hot side flim coeficent, the conductance of SS316 and the cold side coefficient, the hot side dominates and resulting in a low overall HTC. I don't see under such a condition that the HTC in PHE1 will have HTC as high as 9000 W/C/m2. It could be someone in the PHE industry, simply using the load divided by the PHE surface area and LMTD, and then obtaining the HTC. How the area is selected from the begining, again is another mistery.

Unless the manufacturers are cooperating to provide data, the PHE selection and performance data is still a puzzle to me, especially the U or HTC.

 

[rmw]

Sam,

Your vendor should have provided a dimension on his product drawing that shows the compressed dimension of the plate pack.

With that dimension, one can subtract the dimension of the thickness of all the plates (e.g. 61 X 0.5mm = 30.5mm) and take the difference between that and the overall compressed plate pack dimension and divide it by the total number of plate gaps and get close to the dimension.

There are a couple of unknowns here such as the end plate configuration (or lack there of) but that won't skew the number very much. In other words if the calculation results in a dimension of 2.91mm, then the gap is probably a nominal 3mm for sizing purposes.

So, if you can provide the dimension of the compressed plate pack, then zerok should be able to go from there.

Zerok will need the dimension of the distance between the centerlines of the top and bottom ports assuming that the dimensions you gave him are the overall plate width dimensions.

rmw