How Can I Design an Efficient Water Heating System for a High Flow Pool?

[IttyBittyBit]

I want to design a water heating system for a special type of pool. The system is open-loop, where input water at 10 C has to be heated to 15 C, and then enters the pool and is eventually exhausted out of the pool. The flow rate is about 100 litres/second (6000 litres/minute). Simply heating up the flow requires about 2 MW of power. The temperature of the exhausted output stream doesn't matter, so I was thinking of using a heat exchanger to reclaim some of the heat. A few back-of-the-envelope calculatations led me to believe that because of the large flow rate and the pretty low temperature difference, a very huge and costly heat exchanger would be required to achieve any worthwhile heat reclamation. So I'm now thinking of using a heat pump system to reduce total system size and increase efficiency. The idea is an on-site diesel or gas-powered engine running a compressor, where the input stream is heated by the engine, engine exhaust, and heat pump hot end, and the output stream is cooled by the heat pump cold end. Searching on the internet reveals that systems like this are sometimes used for pools, but the flow rate is typically much lower than what I'm seeking and not scalable. How would I go about designing this?

 

[SteamKing]

I want to design a water heating system for a special type of pool. The system is open-loop, where input water at 10 C has to be heated to 15 C, and then enters the pool and is eventually exhausted out of the pool. The flow rate is about 100 litres/second (6000 litres/minute). Simply heating up the flow requires about 2 MW of power. The temperature of the exhausted output stream doesn't matter, so I was thinking of using a heat exchanger to reclaim some of the heat. A few back-of-the-envelope calculatations led me to believe that because of the large flow rate and the pretty low temperature difference, a very huge and costly heat exchanger would be required to achieve any worthwhile heat reclamation. So I'm now thinking of using a heat pump system to reduce total system size and increase efficiency. The idea is an on-site diesel or gas-powered engine running a compressor, where the input stream is heated by the engine, engine exhaust, and heat pump hot end, and the output stream is cooled by the heat pump cold end. Searching on the internet reveals that systems like this are sometimes used for pools, but the flow rate is typically much lower than what I'm seeking and not scalable. How would I go about designing this?

You only need 2 MW of input heating power if you plan to instantly raise the temperature of the incoming 100 L/s of water by 5° C in one second.

You don't mention the filled capacity of the pool, so the total duty of the heating system is unknown. Water is good at retaining heat as long as its container is well-insulated and you can accurately estimate the heat loss from the free surface, which heat loss must be replaced anyway if you hope to keep the temperature of the pool at 15° C.

If we sized kitchen stoves in this manner, we'd need a small blast furnace to raise the temperature of a tea kettle from say 10° C to 100° C in one second.

 

[IttyBittyBit]

I do need 2 MW of heating power. This is a continuous open-loop system. 100 L/s in, 100 L/s out, operating 24/7. I should have made this more clear.

 

[CWatters]

This company say they make heat pumps from 1.5kW to 1MW...
http://www.viessmann.co.uk/en/products/detached_semi-detached_houses/Products/Heat_pumps.htm

 

[jim hardy]

how does so much heat escape the pool? 2mw is 2681 horsepower...

 

[billy_joule]

Have you priced running costs? Running 24-7 is 3x10⁹ litres/year and 6x10¹³ joules/year.
What is the point of this thing? 15C is still too cold for swimming for most... And why is all the water dumped?

 

[CWatters]

how does so much heat escape the pool? 2mw is 2681 horsepower...

He's discharging water and has to heat fresh cold water.

 

[jim hardy]

2 megawatts at 10c/kwh = $200/hr to heat that water on the way in
so
it seems worth trying to recover some heat on the way out with a chiller as he proposed

that's going to be a big chiller...
Is my back of a napkin arithmetic right - almost 3600 tons of refrigeration to cool his flow back down to 10C ?

600 l/sec X 0.0353147 ft³/l = 21.2 ft³sec X 3600 sec /hr X 62.4 lb/ft³ = 4.76 X 10⁶ lb/hr ?
X 9 degF temperature change = 42.8 million BTU/hr
divided by12,000 BTU per hour to the ton = 3570 tons of refrigeration ?

10C is 40F, about the lower limit for absorption coolers
if he has natural gas available that energy cost might beat electricity
not my area of expertise
but here's a catalog
http://www.trane.com/content/dam/Tr...tion-liquid/fuel_drivenabsorptionchillers.pdf

 

[russ_watters]

Welcome to PF!

I agree with the others that the application is quite odd and would question the need to use once-through water. That said, pretty much any standard water cooled HVAC chiller will meet your needs. These temperatures and heat capacity are right in the wheelhouse for what a typical commercial HVAC chiller does.
Try these:
http://www.trane.com/commercial/nor...iller/centrifugal-liquid-cooled-chillers.html
2 MW heating is about 450 Tons-Refrigeration (or search for the European/SI catalogs).

Expect the installation to cost around half a million dollars and cost $400,000 per year to run if it runs all the time.

 

[russ_watters]

Is my back of a napkin arithmetic right - almost 3600 tons of refrigeration

There are 3.413 BTU/hr in a watt and 12,000 BTU/hr per ton. So that's 2,000,000 * 3.413/12,000 = 568 TR.

You did a lot of extra conversions you don't need since you already have the watts (assuming the 2 MW was correctly calculated)

10C is 40F, about the lower limit for absorption coolers
if he has natural gas available that energy cost might beat electricity

Because this is a heating application, you get about 20% more on the heating side than on the cooling side (at least for electric chillers - not quite sure about absorption, but if anything they are less efficient). So if the OP needs a 5C rise on the inlet, he'll get a 4C drop on the outlet (my calculations include that discount on the chiller size).

Also, 10C is 50F -- it's actually pretty warm for a chiller, so would be at a very efficient operating point. But yes, absorption may be a better option if there is fuel available and they don't otherwise need/want the electricity (or use the waste heat from the generator).

 

[jim hardy]

Also, 10C is 50F -- i

Thanks - I'm getting mighty accident prone.

 

[CWatters]

It might be cheaper and easier to clean the water or fix whatever it is that prevents you retaining the water you paid to heat.

 

[jim hardy]

(assuming the 2 MW was correctly calculated)

that's what i wanted to check

i always calculate two or three ways
and try to resolve my inevitable arithmetic mistakes

a liter of water is close to a kilogram
its heat content is 4.187 kJ/kg ? http://www.engineeringtoolbox.com/water-thermal-properties-d_162.html600kg/sec X 4.187 kJ/kgK = 2512.2 kJ/secK
2512.2 kJ/secK X 5K = 12561 kJ/sec = 12.56 megawatts
12,560,000 watts X 3.413BTU/wh = 4,267 X10⁷BTU/h
4,267 X10⁷BTU/h /12000 BTU/tonh = 3572 tons?

I got ~same answer twice

but AHA ! I also made same mistake twice - used wrong flow

The flow rate is about 100 litres/second (6000 litres/minute).

i used 600 l/sec not 100
so i was off by factor of 6

my 3572 tons / 6 = 595 tons
and
my 12.56 megawatts / 6 = 2.09 mw

now we're all in same ballpark

Thanks !

short term memory is the first to go ? Mine's long gone ... can't hardly remember page to keyboard anymore

 

[anorlunda]

The idea is an on-site diesel or gas-powered engine running a compressor, where the input stream is heated by the engine, engine exhaust, and heat pump hot end, and the output stream is cooled by the heat pump cold end.

The simpler the better. Every time you have an energy conversion step, it is necessarily less than 100% efficient.

If you are starting with fossil fuel, then heating the water directly from burning fuel is the simplest method I can think of.

You didn't mention cost. Therefore, let me add my personal favorite (and extremely efficient) form of water heater -- a nuclear reactor.

 

[jim hardy]

If he could cool the exit stream to let's just say 5C
he'd extract most of his heat need from it

what is Carnot limit for heat pump operating between 5C and 15C ?

http://hyperphysics.phy-astr.gsu.edu/hbase/thermo/heaeng.html#c3
do i remember correctly
15C = 288 K
5C=278K

CP = 288/(288-278) 288/10 = 28.8

this is where heat pumps shine, low temperature difference

When used for heating a building with an outside temperature of, for example, 10 °C, a typical air-source heat pump (ASHP) has a COP of 3 to 4, whereas an electrical resistance heater has a COP of 1.0. That is, one joule of electrical energy will cause a resistance heater to produce only one joule of useful heat, while under ideal conditions, one joule of electrical energy can cause a heat pump to move three or four joules of heat from a cooler place to a warmer place. Note that an air source heat pump is more efficient in hotter climates than cooler ones, so when the weather is much warmer the unit will perform with a higher COP (as it has a smaller temperature gap to bridge). When there is a wide temperature differential between the hot and cold reservoirs, the COP is lower (worse). In extreme cold weather the COP will go down to 1.0.

https://en.wikipedia.org/wiki/Heat_pump

If he could actually get a CP of 3

would his 2 mw drop to 2/3 = ~0,7 mw?
At 10 cents a kwh that's $130 an hour !

 

[anorlunda]

this is where heat pumps shine, low temperature difference

You're quite right Jim. I forgot about heat pumps when I said that each step necessarily loses energy.

 

[David_King]

Couldn't most of the heat transfer from hotter exhaust to cooler incoming water happen passively via a heat exchanger? That should be able to recover 85% or more of the heat.

 

[jim hardy]

Couldn't most of the heat transfer from hotter exhaust to cooler incoming water happen passively via a heat exchanger? That should be able to recover 85% or more of the heat.

A counterflow hx sure seems intuitive
i'm not enough of a mechanical engineer to size one for such small terminal difference

 

[insightful]

The "go to" heat exchanger for high flow and close temperature approach is the plate-and-frame type:

http://www.tranter.com/Pages/products/plate-heat-exchangers/description-benefits.aspx