I think I want to measure duty cycle (also called Load Factor, or "LF") on some time period longer than 1 hour. I already have 1-hour duty cycles, putting them into a graph would make the info far easier to read -- but maybe my spreadsheet skills are too basic, I think it would take me a lot of work.
My thinking is that 100% 1-hour duty cycle is not measuring the threshold of sizing. My system which has run 2+ hours without stopping, seems always able to maintain the 78-degree thermostat setpoint (because of my duct qualities, temperature is often 2 degrees cooler in bedrooms and bathrooms). Very typical behavior is running 55 minutes out of each hour, for several hours in a row. I believe this illustrates the system barely making the setpoint, then it would want to restart sooner but the 5-minute off-time is enforced by something in the system (you can tell I am not a pro, with language like that!).
For the interest of the original poster, here are some of my hourly load factors:
Jul 01 9-10pm 73% LF i.e., ran about 44 minutes out of 60.
10-11pm 70% LF
11-12pm 53% LF
12-1am 55% LF
1-2am 58% LF
2-3am 58% LF
I looked at other days and believe July 1 is completely typical (it helps that S.Texas has a lotta typical summer days). It may seem curious to some, but mid-morning is when this system sees lowest LF. I don't know how to explain, but I am reporting the data as I observe it. Will admit possibly I am off by one hour due to daylight savings time, but no really big experimental flaws that I know of. So have a heart and don't call me a "clown" or "Homer Simpson" if my experimental data does not cleanly fit one's expectations. It simply is observed data.
It may help to know that my #1 purpose with these measurements is to provide a non-Manual-J method of finding out whether my system sizing is appropriate, with *my* usage patterns, *my* duct and other flaws as installed, etc. In other words, an observed LF below 100% is incontrovertible proof the system is doing the job I ask of it. Any LF significantly below 90% is proof IMO that system could be replaced with a smaller one and be properly sized. It is not my intention to be controversial, and I intend to have a proper Manual J report made to "prove it" via conventional wisdom before I buy a new AC. But to get original, direct evidence seems pretty darn good to me. That might explain why my data is in a big spreadsheet and maybe not formatted to communicate best.
Best wishes -- P.Student
[Edited by perpetual_student on 08-17-2005 at 10:03 AM]