Window Air Conditioner to Cool an RV

There is has been an interesting discussion on the ProMaster forum on using a window air conditioning unit to cool camper van conversions.  These window AC units are compact, cheap and fairly efficient, so they could be a good solution for cooling camper vans which are typically less than 100 sqft.

One way to install one of these units in a camper van is with the AC unit fully inside the van (say under a bed platform).  Normally these units are installed in a window, and the condenser coil on the AC is outside and a condenser cooling fan picks up outside air from vents in the sides of the unit, forces this air through the condenser heat exchanger, and exhaust it out the back.  If the unit is installed inside the van, then provisions must be made for cooling air for the condenser coil.   Most of these installs use inlet ducts that go through the floor to supply the intake for condenser cooling air, and the condenser outlet vent can either be a rectangular vent cut in the side of the van, or more holes through the floor.



Its unclear how much these floor and sidewall vents impact the performance of the window AC unit.

Quick bottom line:

Its clear from the table below that the degradation in AC performance as you restrict the condenser cooling area can be severe — all the way up to having the AC shut down.

As you restrict condenser cooling air, the condenser temperature goes up, the cooling provided goes down, and the power consumption goes up quite a bit.  If you are trying to get to an AC that will run off batteries, the added power consumption would be a problem.

On this 10K BTU AC, the configuration with four 4 inch condenser inlet ducts and no restrictions at all on the exhaust still increased condenser temperature by 10%, increased power consumption by 9%, and reduced cooling by 2%.

Going to a configuration with two 4 inch inlet ducts and two 4 inch outlet ducts increased condenser temperature by 43%, increased power consumption by 54%, and decreased cooling  by 15%.  This would effectively make an EER 10 AC into an EER 5.5.  This seems pretty dismal to me, and the AC may not even be willing to operate for long periods under these conditions.

It seems like one approach that could be workable would be to place a louver vent panel in the side of the van that is about the same size as the AC’s heat exchanger (16 by 11 inches on this AC) to serve as the AC exhaust vent, and then use as much duct area through the floor as you can get in for the condenser intake.

Update: Added a test (bottom of this page) using a fan to force air through the condenser heat exchanger.  This works quite well — the performance is essentially the same as the bare AC.  See test below.

Test Setup:

The test described below looks at how much these restrictions on the condenser cooling air flow effects performance.  The tests are done on a window AC that I had had lying around.  Its a 10,000 BTU/hr Fedders.  At full cool, the nameplate says it draws 1100 watts. It has an EER of 9.0.  So, its larger than ideal, but it should give an idea of what goes on.  The AC is probably about 12 years old.

I built an outer housing for the AC unit that allows restricting the flow of cooling air to the condenser both on the intake and exhaust side.   The amount of restriction can be adjusted by plugging some of the 4 inch pipe vents.

There are four 4 inch ducts on the sides that can be used to restrict the intake cooling air flowing into the condenser, and there are four 4 inch ducts on the back panel that can be used to restrict condenser cooling air exhaust flow.  The back can be left off entirely to test for no exhaust flow restriction.

There is 2.5 inches of clearance between the added housing and the AC unit on the sides, and 4 inches on the top.

The pipes are nominal 4 inch (actual 3  7/8 inch ID).

There is also an added blower/fan on top of the housing that can be used to provide a forced flow of air into the condenser intake.  To be tested later.


The following five configurations with ever increasing vent area restrictions were tested:

  1. Basic AC unit with no restrictions (as a baseline)
  2. Condenser Intake: four 4 inch diameter ducts
    Condenser Exhaust: no restriction at all — back not installed.
  3. Condenser Intake: four 4 inch diameter ducts
    Condenser Exhaust: four 4 inch diameter ducts
  4. Condenser Intake:  two 4 inch diameter ducts
    Condenser Exhaust: four 4 inch diameter ducts
  5. Condenser Intake: two 4 inch diameter ducts
    Condenser Exhaust: two 4 inch diameter ducts
  6. Condenser Intake: one 4 inch diameter duct
    Condenser Exhaust: one 4 inch diameter duct
    This last config made the AC go into some kind of protective mode and was aborted.


This table below summarizes the results — it shows how much the power consumption, condenser temperature go up and how much cooling capacity goes down for each increase in flow restriction.

Bear in mind that this is a 10K BTU/hr and most of the window ACs used in camper vans are 5K or 6K BTU/hr — so, somewhat less vent area may be needed for the same performance.

Configuration number 1 2 3 4 5
Config: Cond Inlet base AC four 4″ ducts four 4″ ducts two 4″ ducts two 4″ ducts
Config: Cond Outlet base AC full open – no back four 4″ ducts four 4″ ducts two 4″ ducts
     Inlet Area (sqft) 1.125 0.33 0.33 0.16 0.16
    Inlet Area % reduction 0 71% 71% 85% 85%
     Airflow (cfm)(1) 871?(3) 299 264 160 141
     Airflow % reduction 0 ?
     Cond Temperture (F) 112.1 122.8 135.8 147.4 160
     Cond Temperature % increase 0 9.50% 21.10% 31.50% 42.70%
     Power (watts) 1040 1130 1250 1390 1598
     Power increase % 0 9% 20% 33.60% 53.60%
     Evaporator temp drop (F) 36.8 36.1 34.3 31.9 31.2
     Cooling % reduction 0 2% 7% 13.30% 15.20%
1 – airflow velocity calcuated based on air velocity at the 4 inch inlet ducts.
2 – KillAWatt meter tripped and cut off power at this point.  When AC was started without KillAWatt it went into a low cooling mode — overheat protection?s
3  – very approximate.  Measured condenser outlet velocities at four points and averaged.


Parameters measured: power consumption,  evaporator inlet and outlet temperatures, condenser inlet and outlet temperatures, airflow.

Airflow is estimated by measuring air velocities using a Kestrel Anemometer.

The area of the condenser heat exchanger is 1.26 sqft.

The area of the intake grills for the condenser cooling air is 1.125 sqft.

The exhaust grill on the AC is actually larger than the condenser coil so it does not restrict it at all.

Each 4 inch duct (actual ID 3.875 inches) provides 0.082 sqft of area, so 4 of them provide 0.33 sqft of area — an about 70% reduction in area compared to the intake grill on the actual AC.

The plot below is from the logger that was used to log the temperature sensors.  The added numbers (1 throug 5) are for the various duct configurations (see above).

Red dash — Condenser exhaust air temperature (F)
Black solid — Condenser inlet temperature  (F)
Black dash — Evaporator inlet temperature (F)
Green long dash  — Evaporator outlet temperature (F)


I ran each configuration for 15 or 20 minutes to let the AC get to a stable condition, but you can see that the condenser temp is still going slowly up at the end of each test — so, the condenser temps shown in the table are not quite up to what they would be for a long run.

The plot shows the condenser inlet temperature rising toward the end, this was likely bad sensor placement, as the condenser air was just being drawn from the room (as was the evaporator intake air).

I tried to go down to only one 4 inch intake duct and one 4 inch exhaust duct.  Very shortly after I did this, the Kilowatt went up over 1800 watts and started beeping (its only good to 1500 watts), and then tripped and shut off the power.  I took the Kilowatt out and plugged the AC in again.  In a short time it apparently tripped something in the AC — it continued to operate and produced some cool air, but condenser temperature and cooling output dropped.  So, I guess the AC was saying enough is enough!  I turned the AC off, let it cool off overnight, and it seems to be back to normal now.

Test With Inlet Fan on Window AC (much better)

This test looks at using a fan to force air into the condenser heat exchanger inlet.  The idea being to try to achieve about the same airflow as the base AC operating free of any restriction.

Bottom line is that this worked well.  It achieved essentially the same performance as the base AC with no restrictions on the condenser cooling air.

The fan is a Dayton 4WT44A from Grainger.  It is rated 680 cfm delivering to free air .  It appears to achieve about the same volume of cooling air through the condenser heat exchanger as the base AC operating with no restrictions does.  The fan draws 25 watts, so only a small fraction of the 1000 watts the AC itself draws.

Dayton fan

More info on the Dayton fan…

In the test setup, the fan blows air into the space between the cover over the AC and the AC to provide the condenser heat exhanger with cooling air.  The performance of the AC with the cover and inlet fan is nearly the same as the bare AC, so this configuration works quite a bit better than the non-fan driven configurations of the first test (above).

One way to set this up in a van would be to have the fan fit tightly over an opening in the floor such that the fan blows air into the space between the AC unit and the cabinet/cover you have enclosed the AC unit in.  The opening in the floor should be about the same size as the fan (so as not to restrict airflow), and the cabinet over the AC should provide enough space between AC and cabinet for good airflow.  The cabinet needs to be sealed well enough to prevent air leakage into the van.    The condenser exit must be vented out through the side of the van with a vent that is large enough not to restrict the flow — ideally about the same size at the AC condenser heat exhanger.  The outlet vent could also be through the floor, but I think it would be hard to find enough area for a good vent.

Bear in mind that this AC is a 10,000 BTU per hour unit, so if you are using a 5000 BTU/hr unit, the fan and vent areas could be correspondingly smaller

  1. Basic AC unit with no restrictions (as a baseline)
  2. Condenser Intake: Fan pushes are through the condenser heat exchanger.
    Condenser Exhaust: no restriction at all — back not installed.
  3. Back to the Basic AC.


Configuration number 1 2 3
Config: Cond Inlet base AC 9″ inlet fan baseAC
Config: Cond Outlet base AC full open – no back baseAC
     Outlet Area (sqft) 1.26 1.26 1.26
     Outlet Area % reduction 0 0% 0%
     Airflow (cfm)(1) 811 814 888
     Airflow % reduction base 0%
     Cond Temperture (F) 112.7 115.7 114.5
     Cond Temperature % increase 0 2.60% -1.00%
     Power (watts) 1040 1085 1065
     Power increase % 0 4% -2%
     Evaporator temp drop (F) 34.7 35.1 35.6
     Cooling % reduction 0 -1% -1%
(1) airflow is from 4 velocity readings over condenser heat exchanger outlet — pretty approximate.

So, the fan forced cooling of the AC condenser heat exchanger provides the same cooling with no increase in power consumption (other than the fan) as the base AC.

Note that there is a discrepancy between the 680 cfm the fan is rated at and the 815 for so cfm in the table above — this is probably due mostly to the approximate way that I measure the condenser airflow — I just take 4 air velocity readings at the quarter points on the heat exchanger, then average them, and multiply by the heat exhanger area.  So, not so accurate, but it does give a way to compare airflow between the base AC configuration and the fan forced air configuration.   The result is that the airflow is very similar — which is good.

Fan:  note that a lot of fans are rated on how much air they deliver when operating into free air (ie no resistance).  This is not the case for this application — there will be some flow resistance and the fan has be able to produce good flow with this resistance.  As a starting point, if using a 5000 BTU/hr Ac, I’d look for a fan that can deliver of the order of 400 cfm while working against a pressure drop of about 0.1 inches of water.  This is just an educated guess, but maybe better than nothing.  Good fan manufacturers provide a plot that shows airflow delivery vs pressure drop in inches of water.

Plot shows AC temperatures over the test.

Red – condenser heat exchanger exit temperature (F)

Black solid — condenser heat exchanger inlet temperature (F)

Black dash — evaporator coil inlet temperature (F)

Green – evaporator  coil outlet temperature (F)



Any comments, suggestions, ideas … are most welcome.






  1. Gary: Brilliant idea, and your testing is worth a ton of theory. Thanks!

    Coupla questions. In the original version, William Bullivant appears to have a) used a fan with only a quarter of the thru-put of yours; b) provided nominal separation (using the foam baffle) between condenser intake and exhaust, while as far as I can tell you provide no such separation.

    It had not occurred to me before reading your article that the condenser intake and exhaust might not need to be completely separated. Is it true that you don’t separate them at all? And do you have any comments on William’s ability to get away with 150 cfm?

    • Hi EJ,
      I separated the condenser input and output, and it seems to me you should separate them.

      I think hat William was going to do some more testing with temperature measurements, and I’d like to see these.

      It seems to me that providing good ventilation openings or the condenser flow is important for both efficiency (low power consumption), cooling output, and life of the AC. I’d try as hard as you can to provide good condenser cooling flow.


      • Gary: Thank you for your speedy reply.

        However, I think I may have inadvertently said something ambiguous. I understand that the box input and output are separated on the condenser side of the window unit. What I meant to say is that I thought – until I saw your pix and William’s pix – that the *box’s* access to the *unit’s* condenser vents had to be isolated completely from the *unit’s* condenser grille output to the *box’s* output. In other words, I had assumed that intake air into the box had to be forced *in its entirety* through the unit’s condenser side.

        From your (and William’s) descriptions, as far as I can tell, it doesn’t matter that the air coming into the box isn’t *made* to pass through the unit. You are merely concerned that the air in the box is constantly refreshed, and you allow the unit’s own fan to draw that air through the condenser side of the unit.

        As to the cfm of the fan, I am looking at continuous duty fans and their thru-put because I’d like to make any holes in the van as small as possible. But it’s clear from your measurements that getting plenty of air into that box is critical to its working well.


  2. Hi Gary,

    I’m also curious about the relative impact for reductions in inlet area vs discharge area. Reductions in discharge area seem to have a larger impact based on your data above, although it is hard to tell exactly how much larger. I assume that this is partly due to hotter (less dense) air at the discharge, and also possibly due to the fact that the discharge is more “uniform” in the unrestricted configuration? I’m also curious whether a booster fan on the outlet would have a similar benefit if some reduction in outlet area is necessary. Guess I’m going to have to order my air conditioner and start running a few tests!

    Thanks again,

    • Hi Andy,
      I’ve also wondered about placing the fan on the output side as it could reduce the size of the output grill on the side of the van and would allow the AC to sit closer to the floor without the fan under it.
      I don’t see why this would not work. The fan characteristics might change some for “pulling” the air rather than “pushing” it.


  3. Hi Andy,
    That’s a good point.

    Config 2 on the original test has the four 4 inch ducts and no back. The four 4 inch ducts have just about the same area as the net area of the 9 inch fan (that is the 9 inches minus the 3.5 inch hub).

    But, it would be good to know how it would do with just a 9 inch open hole — 63.5 sqi inches compared to the 50.2 sq inches for the four 4 inch ducts. Or, stated another way, how much inlet vent area would it take to give nearly the same performance as the base AC unit. I’ll see if I can run that.


  4. Gary,

    As always, you have done an amazing job analyzing the situation and collecting data. I’m curious if you ran a similar test to the last set, with the fan install, but not operating in order to quantify the benefit of the fan itself. Since the inlet area doesn’t match any of your previous tests, we don’t know for sure what the performance would be with a roughly 9″ diameter opening.

    Keep up the great work. I am looking at installing a window AC in my Promaster and your testing has really helped to inform my decision.

Leave a Reply

Your email address will not be published. Required fields are marked *