Wednesday, August 7, 2002

New A/C System

On 9 May 2002, I had the A/C system checked out to make sure that it would work for the upcoming summer. While it is working correctly, the technician said that it is "running high amps" and will fail under load. 

Since we should hardly need the A/C system in May, but will definitely need it in July thru September. So it seems like a good time to look for a new A/C system before it fails. I checked with 3 different companies. It seems that a good system would be a Trane XL1800, 16 SEER, 5 ton unit. We want to replace our existing system, both heating and cooling. The Trane XL1800 is common, and we can get comparative quotes. 

From 3 different suppliers we got: $8970, $9177, and $7624. An alternative was presented by Strand Brothers. They quoted a Lennox system which was SEER 14.5 for $6187. Instead of Freon, it uses 410A as a refrigerant. Plus it runs at variable speed, trying to minimize how much it has to do to reach a target temperature. 


But before we do that, we check with the City and they suggest we have our ducts checked. The duct work for our heating and A/C system run thru the attic. We've had all the ducts replaced once before, in April 1996, but the City believes that a lot of conditioned air is lost due to leaky ducts, so they are supporting a duct leakage analysis. We had our ducts tested on 3 June 2002. The duct leakage analysis showed that we both had duct leakage, but possibly more important, not enough air flow to properly cool most of the rooms. So we need the ducts sealed, but also modified to get the right amount of air, at the right temperature to all the rooms in the house. 

We contracted with Strand for both pieces of work. First, replace the existing heating and cooling system with a new Lennox system (both a gas furnace and A/C system), and second, modify the duct work to adjust the air flow and seal them. The contract, 8 June 2002, was $6187 for the replacement system and $2625 for the ductwork, for a total of $8712. 

I suffer from allergies, and so I would like the air conditioning system to also do the best it can to clean and filter the air. So we added onto this total $650 for a Honeywell Electronic Air Filter, and $395 for an ultra-violet light. The UV light goes anywhere in the air flow from the return air chamber to the top of the system where it branches for the various parts of the house and is supposed to help kill viruses and bacteria. 

We put $5000 down on 20 June 2002. The heating and A/C system was installed promptly and with little problem. The main issue was that the electronic air filter, installed just above the return air chamber raised the fan, furnace and evaporator coil. This makes it difficult to put the evaporator coil in place, and so it had to be turned 90 degrees from what was expected. 

But the duct work was more difficult; much more difficult. The ducts themselves were replaced and rebuilt according to plan, but they did not work as designed. The objective was to increase air flow to each of the rooms to be plus or minus 10% of what would be desired for that size of room. There were 18 registers. 


A spreadsheet was used to identify each register, the desired airflow, and the measured airflow. 

My idea was to have a concrete, objective criteria that defined if the work was "good enough". This seemed to cause significant problems. Strand expected to measure the current situation in order to establish that I needed work done, then design a solution to fix the situation, do the work, and declare it done. They did not normally go back and re-measure things to see if things actually improved. When we did, we found that some registers had improved, but others had not; some actually got worse. Before the work was done, we had 69% of the desired airflow; afterwards we were at 85%.  Better, but not what they promised.

The reduced air flow resulted in some additional changes to the duct work, and a larger evaporator coil, but still did not meet the goal. This got management involved, with different air filters and different measuring devices. The result seemed to be that first, the measuring devices were not very accurate and second, the measuring devices were not even repeatable. 

The measuring devices fit over a register and measured the amount of air coming out of it. Different instances of the measuring device gave different numbers. The same device would also give different results. Part of the problem was the Lennox system itself. It has a variable speed fan, and is reluctant to "blow at full speed". The result of this is that it is difficult to get the fan to run at full speed, or even a constant speed. Lennox apparently did not build into its controls a "test" mode which would just turn the fan on, full speed. 

Eventually, we got things done as well as they were going to be, but it was not easy. We never got the ultraviolet light, and the reusable electrostatic filters were deemed too restrictive to air flow, so we use disposable HEPA filters instead. 

One of the better features of this system turns out to be the warranty. The warranty is for 10 years parts and labor. 


This has turned out to be critical:

  • 23 May 2005 -- the system doesn't seem to be cooling. I have an A/C pocket thermometer, and it shows the air coming out of the closest register to be just barely cooler than the air going in to the return air closet. Turns out there is no coolant; the 410A coolant has leaked out. They put in more, but come back a couple days later and find a leak in the coil, so they replace the coil and recharge the system. This takes about 2 weeks altogether.
  • 1 July 2006 -- the compressor will not run. A bad "Dual Run Capacitor" is replaced.
  • 10 July 2008 -- internal electronics short out. The module controlling the blower motor is replaced.

Each time, Strand came out, found and fixed the problem. No charge. This should continue until 27 June 2012.

But Strand was none too happy about it.  They sent a letter in October 2008 saying that the system was no longer under warranty.  I sent back a copy of the original warranty, showing that it was still in effect until 2012. 

But the arguing ended up being unnecessary.  At this point, the system stopped breaking and just kept running for the next 10 to 12 years, when it was replaced by a Trane.