When you do a heating and cooling load calculation using ACCA Manual J, it’s important to be as accurate as possible with the inputs. A little oversizing isn’t terrible but I’ve seen systems that are more than twice the size they need to be. And I’ve seen similarly oversized equipment recommendations based on inaccurate Manual J’s. The truth, however, is that even if you’re as accurate as possible with your inputs, your loads will still come out a little on the high side, probably ten to fifteen percent.
So what kind of loads should you expect? The old rules of thumb do not work, so if your load is coming in at 500 or 600 square feet per ton, it’s probably either a below-code, inefficient house or you’ve made some mistakes. We do a lot of load calculations and I put together a graph of 40 typical new construction projects. The average home came in at 1,431 sf/ton. (Read the article for more details.) Obviously existing homes can be worse and there’s a lot more uncertainty in the inputs.
If you’re doing a Manual J load calculation, though, here are some of the most common mistakes I see.
- Indoor design temperatures. The standard indoor temperatures are 70° F for heating and 75° F for cooling (with 50% relative humidity).
- Outdoor design temperatures. The outdoor design temperatures depend on the location. To find what should be entered, just look Table 1A in Manual J. You also can go to this page on the International Code Council's website. If the entries in your calculation are off by a couple of degrees, it's not a big deal. If they're off by 5 degrees, you should correct it. (Read more about design temperatures here.)
- Areas. Having the wrong surface areas for the various floors, walls, ceilings, windows, and doors can make a big difference . This is especially true for parts of the building enclosure that have worse specifications, like windows. A code-built house in IECC climate zone 3, for example, has windows that are about R-3 whereas the walls will be R-13. Entering too much window area is a way to inflate the load. Entering too much of any of the areas likewise inflates the load. It can be a lot of work to check all the areas, but if you suspect errors and can't find them elsewhere, get out your calcuator and do it.
- R-values and U-values. Check the entries for the floors, walls, ceilings, and floors to ensure you put in the correct R-values (for insulation) and U-values (for assemblies, like windows).
- Number of occupants. A common mistake that results in a larger cooling load is adding extra occupants. The rule here is that the number of occupants should equal the number of bedrooms plus one. If you put 23 people in a 5 bedroom house (Yes, I really saw that!), you’re adding unnecessary load. At 230 BTU/hr sensible and 200 BTU/hr latent, those 17 extra occupants added more than a half ton of cooling load.
- Infiltration. Did you use a simplified input method? If it’s for a new house that meets a code that requires 7 air changes per hour at 50 Pascals (ACH50) or better, the entry should be “”tight” or “semi-tight.” Better would be to use an actual blower door number. For example, if your code requires 3 ACH50, enter that into the calculation. If the house is going for Passive House certification, enter 0.6 ACH50 or 0.05 cfm50/sf of enclosure.
- Orientation. The software gives the designer the option of using worst case for the orientation. Your load calculation should have the correct orientation or you'll end up with extra load in your reports.
- Duct location. If the ducts are in conditioned space or in an encapsulated attic or crawl space, make sure that gets factored in properly. Doing the load calculations for ducts in an unconditioned attic will result in excess load.
Contractors doing these load calculations often feel compelled to stretch a little bit here and a little bit there. Each little bit doesn't affect the overall load that much but by the time you add them all up, you may be looking at putting in a 4 ton air conditioner where 2.5 tons could work. But here's the thing: Even when you're as stingy as possible with things that add load, you still end up oversized by ten to fifteen percent. So there's no need to add extra load.