What BTU rated heat ex-changer

[RegSass]

Maybe a more direct question was called for, like what BTU rating is needed for my solar hot water system heat ex-changer. I have 12 4x8 solar water heating panels and 1600 gallon storage @ appx 10 gpm. flow rate in closed loop and what configuration will meet that rating.

[J.P.M.]

Maybe a more direct question was called for, like what BTU rating is needed for my solar hot water system heat ex-changer. I have 12 4x8 solar water heating panels and 1600 gallon storage @ appx 10 gpm. flow rate in closed loop and what configuration will meet that rating.

As I wrote, such a HX would probably be designed to handle the max. output of the solar array. Also, as I wrote, or at least alluded, that is easier said than done.

The first 7-8 years of my engineering career were spent on thermal and mechanical design of heat transfer equipment and systems involving mostly heat exchangers of various types, and I was never far from that end of things as time, and I, moved on.

In a general sense, HX sizing involves the search for an overall heat transfer coefficient. Type of HX, materials, fluid mechanics, heat transfer, Thermodynamics and lots of other things play a part and the whole process is quite iterative and interactive.

The min. info. for a SWAG on HX size would start with: Assuming 50/50 glycol/H2O for the working fluid, min/max. temp. of the fluid from the solar collectors (BTW, a flat plate thermal solar collector can be thought of as a specialized type of HX), the min./max. tank temp. likely to occur as well as any off design conditions reasonably possible/likely, and how the tank fluid (assumed to be H2O) will be in contact with the other side of the HX (HX immersed in tank, tank fluid pumped through HX, etc.). Basically All the temp. and flow rates the HX is expected to see under a "normal" range of operating conditions and any off design conditions that might reasonably (or unreasonably) be expected.

Since applications of this type can have operating conditions much wider than many/most industrial heat transfer applications, the use of one design condition is not as safe a way to go, and therefore some latitude in sizing is often warranted, keeping in mind however, that in HX design, oversizing can be as bad or worse than undersizing for a bunch of reasons.

Many years ago, a guy by the name of Frank DeWinter came up with a closed form solution for the performance penalty for a HX in a collector loop (see Duffie & Beckman). That was not unlike the solution for a batch heating/cooling application using and external HX. His work involved using a single pass shell & tube HX. I believe I extended that solution to multipass shell & tube configurations. More of an academic exercise than of much practical use so I've kept it to myself. Long, boring story, but I had fun.

[LucMan]

Assuming 1K btu per square ft per day production for your 12 panels=384000 btu per day÷6 hrs= 64, 000 btu per hr.
There's your starting point.

[LucMan]

Assuming 1K btu per square ft per day production for your 12 panels=384000 btu per day÷6 hrs= 64, 000 btu per hr. Generic answer.
There's your starting point. Now adjust for your geographic location, and winter sun exposure time.

[J.P.M.]

All in the spirit of general discussion and an attempt at a sketch of the Shell & Tube (S & T) design process and and not meant for design:

I'd use an hourly max. output or shorter basis for HX sizing purposes.

Reason: Assuming a constant flowrate, using max. daily ave. output will result in an undersized HX for those portions of a clear day when output may be the highest, say between 1000 and 1400 hrs. solar time. Everything will perform worse or at least underperform when the solar resource is at its maximum.

That doesn't sound like a good way to start a design.

As for array thermal gain, a REAL SWAG at max. output for a reasonably efficient thermal system running at, say, 50% thermal eff, including piping losses to the HX, might run something like (300 BTU/ft.^2*hr)*(.50)* ~ 150 BTU/hr*ft.^2 to a HX.

That would require a duty of ~ 150*384 ~ 57,600 BTU/hr. for the HX.

A 10 GPM flow of 50% Eth. Gly. through the array: 57,600/((10)(500.7)*(1.06)*(.85)) ~ 12.8 deg. F. rise., probably not too bad for the application. For a lot of reasons, I'd try for a series/parallel arrangement that would achieve a turbulent or at least transitional flow through the collector array.

For optimum heat transfer in the HX, and because of the transport properties of the fluids involved, if this were a shell & tube HX application, the glycol would be best placed on the shellside w/the storage flowing on the tubeside.

Now, assuming (again, for talking/illustrative purposes only) a floor (radiant) heating system that can use, say, a 115 deg. F ave .fluid temp. (pretty low - I know, so some size adjust. of that portion of the heating system might be a consideration), but maybe and because the tank temp. may run higher in warm weather due to lack of required draw, the tubeside fluid (the storage tank water) will operate a bit higher and thus over a range of temps., as will the shellside glycol. That makes setting temps. for design purposes a bit of a challenge for the HX design, but if the tank is hot(ter), the collector may well be operating hotter because the weather may be warmer. So, a 1st cut may be to SWAG a glygol inlet temp. to the HX of, say, 130 F. and a tank water inlet to the HX of say 110 F. (and adjust the inlet as below) with, for now, and for simplicity, a 10 GPM flow for each fluid.

Next, find the outlet temps. from the HX for each fluid. For the glycol: a 12.8 deg. drop to bal. the rise through the collector. So, 130-12.8 = 117.2 deg. F. @ the HX outlet.

For the tank water : ~ 57,600/((10)*(500.7)) ~ 11.5 deg. F. temp. rise.

So, for the 115 F, assumed ave. tank temp. the inlet to the HX will be 115 - (11.5/2) ~ 109.3 F. and the outlet (return) to the load will be 109.3+11.5 ~ 120.8 F.

All that, with some magic, gives something called a Log Mean Temp. Difference (a sort of average HX temp.) through the HX of about 8.5 deg. F.

Next, a shortcut: As I wrote, HX design is initially about estimating a heat transfer coeff. (called "U"). For now and for this forum format, l'll only say calc'ing (estimating) "U" is the black art part of HX design, and for now say that using shell & tube design experience, a decent, slightly conservative 1st appox. for U might be ~ 250 BTU/(hr.*ft^2* deg. F.).

Then, given the duty: 57,600 BTU/hr., a "U" of about 250 BTU/hr.*ft^2 *deg., and a temp. dif. of 8.5 deg. F, a 1st cut, approx. HX surface area can be estimated from Duty = "Q" = 57,600 BTU/hr. = (U)* AREA) * temp. diff -->> Area = Q/((U)*Temp. diff.)

--->>> Required area = 57,600/((250)*(8.5)) = ~ 27 ft.^2.

If, using, say 3/8" O.D. tubes (a common size for smaller, non severe duties), that 27 ft.^2 will require about 275 lineal feet of tubing. As a rough 1st cut, that translates into an 8 " dia. HX with ~ 72 ea., 3/8" tubes that are 48" long.

With that 8" shell dia., for best fluid mechanical and heat transfer performance, a multipass HX with some "tight "baffle spacing would be best. Unfortunately, that multipass feature will degrade the 8.5 deg. F temp. diff. fair amount, probably adding about 20% or so the required area and thus the required length.

And that's the simplified version.

I'd look for an atmos. vaporizer and replace in kind.

Take what you want of the above, Scrap the rest.