10 replies to this topic

[Stefano]

Dear members, has anyone of you experienced an heat loss calculation with Hysys Dynamic?
I tried to do this simulation setting the option "detailded heat loss" in Hysys Dynamic (Ver. 7.2) but I'm not very confident with the result. In my opinion the heat loss is very low!
My vessel is horizontal, elliptical heads, 2m diameters, 7.3m lenght, full of water, operating @ 0 barg. I need to maintain a temperature of 40°C with a minimum external temperature of -2.5°C.
Thickness of 11mm (carbon steel), no insulation and installation in pit (no wind effect). From simulation the result is 4 kW.
Any opinion about this?
Thanks a lot

[kkala]

Let us try to quick check heat loss, through the heat transfer coefficient.
Total vessel area S = 2*π*7.3+8.7 = 54.6 m2
Temperature difference Δt = 42.5 oC
Estimated heat loss Q = 4 kW = 3440 kcal/h
Corresponding U = Q/S/Δt = 1.5 kcal/m2/h/oC.
This overall heat transfer coefficient looks low, compared to e.g. 5 kcal/m2/h/oC. Look at http://www.cheresou...insulated-pipe , post No 11 and on.
Nevertheless mentioned overall heat transfer coefficients (Us) in above link are valid for a slight wind, and wind velocity has a strong influence on U.
Let us try to check U under natural convection (wind velocity=0).
G. G. Brown, "Unit Operations", Wiley 1950, Chapter 29, Natural Convection, quotes following heat transfer coefficient (Btu/ft2/h/oF) for vertical plates over 3 ft high: h=0.3 (Δt)^0.25, Δt = temperature difference between air and plate surface (oF).
Assuming this formula applicable (better proposal welcomed), h=0.3*76.5^0.25 = 0.89 Btu/ft2/h/oF = 4.3 kcal/m2/h/oC. This has supposed no heat transfer resistance in water or metal side, actually Δt is lower and h smaller, but resistance in air side must be controlling and overall U cannot be much smaller than 4.3 kcal/m2/h/oC. It could not reach 1.5 kcal/m2/h/oC.
I imagine that pit configuration is not considered in Hysys, the vessel is assumed in ambient air of -2.5 oC around it (with no wind).
Conclusion: According to the above, heat loss of 4 kW (through Hysys Dynamic) is judged underestimated.

Edited by kkala, 06 April 2012 - 12:39 PM.

[Stefano]

kkala, thanks for your calculation: it confirms my idea. The only comment about it is that I'm considering only the wet surface that is, more or less, 46 m2; but I didn't mentioned the liquid level in my previous post. The Overall Heat transfer coefficient calculated by Hysys is about 6.9 kJ/m2h°C.
However I also found some sources where the Overall Heat transfer Coefficient in still air is given at 0.2 Btu/hr ft2 °C that is about 4.1 kJ/m2h°C.

[Shivshankar]

Stefano,

Have a look at this PDF attached and let me know if this can help you.

Regards
Shivshankar

Attached Files

•  2008.pdf   28.54KB   424 downloads

[breizh]

I've attached a chart to support , may be too simple .

Breizh

[Stefano]

@Breizh: before to post my question I used your graph found in an old post, obtaining an heat loss of 27kW (uninsulated tank): this result has reinforced my doubts about software result

@Shivshankar: the theory is critical to the engineering calculation, but in this case the result is strongly influenced by the right choice of the proper coefficients. Moreover my idea was to have a rough estimation of the right value without spending a lot of time in detailed calculation.

Edited by Stefano, 10 April 2012 - 04:32 AM.

[Shivshankar]

Stefano,

Can you put your worked calculations here so that members will have more idea about your work.

Regards
Shivshankar

[kkala]

My understanding (full of liquid) had been about 100% full vessel, but never mind, the issue does not seem to depend on it. 46m2 with 6.9 kJ/m2/h/oC results in 3.75 KW ~ 4 kW.
Hysys gives 1.65 kcal/m2/h/oC, "other source" 1.0 kcal/m2/h/oC (0.2 Btu/ft2/h/oF). This fair agreement is an indication of correct result, if the "other source" is reliable, minimizing possibility of error in treating / inserting data into Hysys.
Sources from the posts give higher overall heat transfer coefficients. I think that "natural convenction", even in a pit, is rather theoretical. Once I was measuring surface temperatures of a rotary drier (~60 oC) in a closed building (fertilizers), yet I felt several air currents from distant slops or openings. So it is not bad to be reasonably in the conservative side.
Comments / data from others would be welcomed.

Edited by kkala, 10 April 2012 - 06:18 AM.

[Stefano]

Given the uncertainty of data involved I agree with you that the best position is to assume a conservative value..

Now, just for feed the discussion, I found in my "Mechanical Engineering Manual" (Ed. Hoepli 2002) a formula valid for horizontal cilindrical surfaces in natural convective contact with air

Constraint: (Gr*Pr)<10^9

h=1.32*(dT/D)^0.25
dT=42.5 K
D=2 m
h=2.83 W/m2°C

Assuming the other terms of equation negligible I can suppose U equal to h (1/U=1/h1+dx/K+1/h2)
So Heat Loss H=2.83*46*42.5=5530 W=5.53kW

[kkala]

1. According to above calculation, h=2.83 W/m2/oK = 2.4 kcal/m2/h/oC (post No 9), versus 4.3 kcal/m2/ h/oC by Brown (post 2). Brown also presents the equation of Mechanical Engineering Manual in English units, h=0.27*(dT/D)^0.25 Btu/ft2/h/oF, dT in oF and D in ft, noting that this equation is to be used cautiously, it may be unsound at diameters larger than 1 ft. Probably this has been proved invalid (Brown's book was published in 1950), but had better be verified.
2. What is conservative in this case? Higher heat loss is understood to be so.
3. Referring to post No 6, chart indicates U a bit higher than 12 kcal / m2/h/oC (at conditions) for stil air. It concerns tanks with two flat ends, since their surface area is πD²/2. It is conservative. Chart by Armstrong, http://www.armstrong...s/pdf/AH630.pdf '> http://www.armstrong...s/pdf/AH630.pdf , indicates heat losses corresponding to U ~ 8.3 kcal/m2/h/oC under same conditions.
4. 2008.pdf attached to post No 4 indicates 4.3 kcal/m2/h/oC (6.20, natural convection?), based on the lowest value of given range 5-30 W/m2/oK.

Edited by kkala, 10 April 2012 - 04:23 PM.

[Stefano]

Thanks to all who partecipated to the discussion: it was very interesting. About the value to put in the electrical heater data sheet, I'm thinking to set it from a minimum of 10W/m2°C to a maximum of 20 W/m2°C, taking into account all your valuable posts.