Shelter Heat Load Calculations

[SmileyC]

Let me start by saying that I am an Electronic Engineer, and whilst I have enjoyed exercising the grey matter perusing my first year thermodynamics books, I have doubts over the voracity of the following logic and assumptions. Hence the post to this forum.

The problem:

I am trying to determine the heat load associated with equipment installed within an airconditioned ISO shelter. I see three main sources of heat: 1. Equipment, 2. Heat transfer from high external ambient temps (30 deg C+) through shelter walls/ceiling/floor, 3. Direct sunlight
on shelter externals. At this stage I do not need to be able to separate out each component.

The system is essentially closed with no significant transfer of air inside to out.

Can I use the rate of change of temperature (With A/C cooling off and fan on only) to determine the heat load?

I have tried some calculations using:

Q=m*c*(Tf-Ts) joules - (1).

m = mass of air (kg) (Actual - 30.3kg)

I calculated mass of air using density of 1.2kg/m3*air volume. Volume of air equals shelter internal volume minus a % (20%) for the physical volume of the installed equipment.

c = Specific Heat of Air 1010 J/Kg.K

Tf(Final) and Ts(Start) in Celsius. (Actual Tf = 23 C, Ts = 20 C)

Now my understanding is that Eq (1) is a measure of the total energy used to change the temp and is not referenced to time.

Using:

Watts = Joules/sec - (2)

I can introduce a time relationship. Here is where I am unsure of my logic.

From (1) I get a figure for Q of 91,815 Joules.

Converting to Watts, I have assumed that it will take 91,815 Watts of power to produce the temperature change in 1 second. Is this correct?

If the time taken for the temp to change is 3 minutes (180 sec) the power required would be:

91815/180 = 510 Watts.

Is this a true, albeit rough, reflection of the overall heat load experienced in the shelter?

Initially, I doubted my logic because anecdotal evidence shows that a temp change of 3 deg C in 3 mins is what the shelter is experiencing during the off cycle of the A/C compressor. The figures I have for equipment heat load alone are significantly higher than 510 watts.

What I have described above relates to heat load. I assume that I could apply the same logic and measure the time taken to decrease the temperature and get an indication of cooling power of the A/C system. Am I correct?

In terms of measurement I intend to measure air temp, using a data logged digital thermometer (1 sec Samples) located at the return vent of the A/C system adjacent to the internal A/C thermostat.

Are there any fundamental problems with my logic?

Thanks in advance.

 

[IRstuff]

The biggest problem is that you've effectively assumed that all the heat can get to where it actually changes the air temperature.

BUT, the only way all the air can change temperature is through natural convection and forced convection from the fan.

TTFN

 

[quark]

Yes, you can do it that way. The concept is not totally wrong. But for checking cooling capacity I would suggest to use the formula Q = 4.5 x cfm x (H1-H2) where H1 is enthalpy of entering air (at supply grill) and H2 = enthalpy of leaving air (at return grill). For this you have to know wet bulb temperatures and a psychrometric chart.

Regards,

 

[samienela]

Shall we assume that the fan you mentioned is the one belonging to the air conditioner? If so, your logic looks fine provided that you have started your timing *after* the evaporator coils of the air conditioner have finished warming up to room temperature. Timing the temperature rise in only the first three minutes after the AC shuts off may bias your answer low because of the residual cooling from the AC. If this is the case I suggest repeating the experiment but for a longer period and recording the temp. every 30 seconds for, say, 6-8 minutes with the AC off. Plot temp vs. time and redo the calculations for only the linear part of the curve.

 

[SmileyC]

Thank you to IRstuff, quark and samienela for your responses. I am pleased to see that I was not too far off the mark. I will be taking advice from each of your posts when I set up the test.

Regards

 

[davefitz]

It might work, but you may need to also consider the change in temperature ( and corresponding M and Cp) of the masses in the compartment, if they are giving off heat to the compartment.

For example , if a tnansformer or motor is giving off heat, that component is runnning hotter than the compartment, and if the compartment air temp is increasing ,then the component temp must also increase in order to give off the same amount of heat.

There are other methods- I will forward the references tomorrow. There is an old GE paper that provides guidelines for estimating heat given off by various electical components, and there is also the ASHREA procedure for calculating heat transmitted to buildings as a function of loaction and time of day.

One trick - the solar heat gain can be reduced by using standard aluminum lagging - it has an emmissivity of only 0.15-

 

[davefitz]

The GE references for the heat emmitted by electrical equipment in a shelter:

William MacDonald et al "Energy Losses in Electrical Power Systems" , IEEE trans ind appl IA-24 no 4 May 85 pp 803-820

Also for environmental heat loads on blds, see Ronald Howell "Principles of Heating Ventilating and Air Conditioning , 1998 by ASHREA , Atlanta Ga USA
<www. ashrea.org>