Air conditioner coolant supply line to attic

[Themp]

I have a 34 year old Trane air conditioner that needs to be replaced. The coolant line comes into the basement and then runs to the attic through the first and second floor walls. It is a R-22 system. The contractors I had come in want to cut the line in the basement and attic, flush it with a cleaner and then re-use. I am leaning to opening up the walls and replace the complete line. I would actually do the sheet rock pull down and have it ready for the contractor. Any thoughts on this as the contractors say they do this(re-use line) all the time?

 

[Dana]

You can clean it up refrigerant lines from R22 systems and re-use with R410A, but it's usually not "worth it". Abandoning or pulling it and using new refrigerant line is usually more appropriate, cheaper, and lower risk.

Equipment replacement time is an opportunity moment to get the sizing right, and to look at all aspects of the system. Almost all equipment that was installed in the 1980s was 2-4x oversized for the actual loads, which cuts into both comfort and as-used efficiency. Ideally you would be sizing it based on a third party's (not an HVAC contractor's) Manual-J load calculation, but that's the exception rather than the rule. Most contractors would just replace what was there with something of comparable tonnage and call it a day, and that's usually a mistake. Others would use some rule of thumb such as "a ton per 500 square feet of living space", that too is nearly universally a mistake. Cooling (and heating) loads are much lower than most people (and most HVAC contractors) think, and oversizing "just to be sure you're covered" costs you a bit more cash up front, and even more in efficiency losses over time, but more importantly, you take a hit in comfort.

A right sized system would run nearly constantly during the warmest days of summer, seeming like it's struggling, but still keeping up, or almost keeping up, maybe losing ground a degree or two for a few ours on absolute worst days, well above your 1% outside design temp (= 90F, in Raleigh NC). With the longer duty cycles the conditioned air stays pretty dry, usually under 50% RH even on days when it's cycling.

A 2x oversized system always keeps up, but duty cycles on/off even during the warmest days.

A 4x oversized system is big enough to leave it off all day, come home to a 90F house at 5PM and still chill the place to 70F by 6PM, which is great between 6-8 PM, but does at best a so-so / mediocre job of managing the indoor air's humidity levels.

Which best describes the behavior of your system?

Almost all air conditioning & heating installed in an attic, above the insulation creating necessary holes in the pressure & thermal envelope of the house is a mistake, increasing the effective load and ruining the potential efficiency. Yet this is the most common configuration for air conditioning systems, adding a half-ton or more to the direct load, and sometimes even more from air-handler driven air infiltration due to duct imbalances, duct leakage, and air leakage at the attic floor plane.

Often replacing an air conditioning system old enough to have voted in the '92 election with a right-sized heat pump (rather than AC only) is the right thing to do, depending on the age / condition / efficiency of the heating system.

So, it's up to you how much analysis you want to get into (either here, or paying an energy-nerd for a site visit), but now is the right time to figure it all out. Whatever you install is likely to be there for at least 20 years.

 

[Sponsor]

 

[Themp]

Dana, thanks for the great response. I will replace the complete piping. Our current A/C units and flat coil are 3 tons upstairs and down. With 100,000 BTU furnaces 80% efficiency. The A/C units are GE BTD736A 'Climatuff' which are labeled Trane. In 34 years I have never had to do anything to them other than replace the capacitors and add a hard start capacitor to the upstairs unit as it would blow the fuse at startup.

You are correct they are oversized as my wife plays games with the upstairs unit during the hot days of summer. Has to reduce the temperature to 77 degrees and then back to 78 degrees as it gets to cold. Plus we have a playroom that gets direct sunlight in the afternoon and that room gets quiet warm.

At this point we are retired and not sure how long we are going to stay in the house. So, the dilemma is to go with a cheaper unit or stick with Trane. Again, I know the installer is key here. Trane seems to have more control over who they make as dealers. Carrier I do not know much about. Goodman is popular now in my neighborhood. The reason for doing a replacement is that we had some health issues this summer and having an A/C failure during that period would have been painful. So, getting a new HVAC system would remove this worry down the road.

 

[Dana]

How big is this house? You have two 3-tonners, one upstairs the other downstairs, with a 100,000 BTU/hr furnace on each? Unless it's gia-normous house it's all VERY oversized.

Typical real cooling loads for upper floors is about a ton per 1000' of conditioned space, and for first floors with a conditioned floor above it, about a ton per 1200-1500' (depending on how much unshaded west facing window area there is.) A typical 2400' 2-story would come in at around 2 tons for the whole shebang. If you measure the duty cycle on the compressor or condenser blower on a 88-92F day you'd be able to get a good handle on the true cooling load, which would allow you to replace it with something more appropriately sized.

A duty-cycle based load measurement includes all of the parasitic duct gains, and will probably be higher than if you were bringing all the ducts inside, but if you're a short-termer a complete reworking of the ducts to ring them inside would not necessarily be cost effective, but if the playroom is chronically too warm you may want to tweak the balance a bit.

Heating loads for R11-R12 insulated 2x4 framed houses with clear-glass double panes and R19-R25 in the attic typically come in at about 10-12 BTU/hr per square foot of conditioned space @ Raleigh's +20F 99% outside design temp, but designing for 1.4x that much heater is the ASHRAE recommendation to fully cover the 25 year low temperature and have reasonable recovery times. Even if the 100,000 BTU/hr is the input, you'd have 80,000 BTU/hr of output or , or 160,000 BTU/hr of total furnace output out of the pair of them. That's great if your house 10,000 square feet or something, but I suspect it's a wee bit smaller than that. (For reference, my antique 2x4 framed 2400' 1.5 story house with over 1500' of basement comes in at about 27-28,000 BTU/hr @ +20F.)

Your true heat loads are likely to be well within the output of heat pumps appropriately sized for the cooling loads, without needing to engage the resistance heating strip backup. Whether it's cheaper to heat with heat pumps rather than with gas depends a lot on your local electric & gas rates and the HSPF efficiency of the heat pump, but it often is.

 

[Themp]

Total square foot is 3000. With it split evenly on a story and half. Heat pump is an interesting idea as right now the two gas furnaces are vented to a chimney flue and the HVAC folks that came by want to not use them. Put a hole in the roof and a higher efficiency furnace in the crawl space so they can just use PVC to the outside. I wanted to sleeve both flues and re-use and stick with 80% efficiency.

 

[Dana]

If you have some mid-winter gas bills with the EXACT meter reading dates and quantities and a ZIP code (for more refined weather data) it's pretty easy to put an upper bound on your 99% heat load based on fuel use, which would give you a better idea as to whether a heat pump sized to the cooling load would cover your actual heat load with any margin. The methodology is pretty straightforward, as outlined in this bit o' bloggery. If you like, I can walk you through those numbers here if you care to share the informaton.

With 1500' of space upstairs and the parasitic aspects of ducts & air handler being in the attic you may be looking at as much as 2-tons of cooling load up there, but it might still be in the 1.5 ton range. Some duty cycle monitoring on the existing unit would tell. On the first floor it's unlikely you have 1.5 tons of cooling load, but that's the smallest end from most split-system AC & heat pump vendors (until you get into modulating mini-splits). Some series don't go lower than 2 tons. (The Trane XR14 and XR15 heat pumps come as small as 1.5 tons, but their higher efficiency series start at 2. But 2 ton 2-stagers can run with lower than 1.5 tons of output.)

 

[Themp]

Dana, I read your blog posts and some other links from that site and in the end my thoughts are that a 'smart' thermostat that could be re-used on the upgrade and could determine how much energy your system is using. My gas bill would not work since a gas water heater is in the mix. But a thermostat for up and down systems that could be programmed for your zip code could over time determine the energy usage of the system. My current thermostat tells me hours that the blower ran for filter replacement and that could be used.

Of the three estimates that I got for the replacement only one had a 2.5 ton A/C unit on the first floor. The others replaced with 3 ton based on what we had. The estimate for the 2.5 ton on the first floor was the most detailed as the salesman used some program on his laptop. But I did not get any analysis results to look at. For furnaces for the detailed one had 80k up and 60k BTU down. The other two did not specify the BTU.

The other problem is that the ductwork needs to be replaced. The flexible air ducts are easy but in this house the original ductwork included a long metal duct running the length of the attic and crawl space. Off this metal duct the flexible ducts run to the floor or ceiling vents. This metal duct is poorly insulated on the inside and is not taped at the joints. The three estimates did not want to mess with the metal duct. Just replace the flexible with higher R-value ducts. The metal duct does have dampers in them so that air can be diverted to different parts of the house. But in the attic I have it wide open to the playroom.

 

[Dana]

A gas water meter would only insert a low single-digits error if the calculation was done only on mid & late winter gas billing. Only if your family has a 200 gallon spa bath that gets used daily or some other extreme hot water use application (are you heating swimming pool all winter?) would the water heating error exceed that of other errors such as solar gain, and the actual vs. nameplate combustion efficiency. The error from the water heater is an increase in the estimated heat load, but during the middle of the heating season it's a much smaller fraction of the total fuel use, whereas during the summer water heating dominates. The error from solar gains are smaller in winter than in spring or fall, and the net effect on the fuel-use calculation is to lower the estimated heat load. These errors tend to balance each other out, but are only a second-order effects in the mid-December to early March time frame.

There is no way that you could have anywhere near (80K + 60K=) 140K of heat load, unless you sleep with the windows open on the coldest night of the year, and a fuel-used derived calculation would prove that. (Even in New England I've yet to see a 3000' house with a heat load that high, at much lower outside design temperatures.) Roughly 1/3 that number is a more likely number, or about 45K- less if your house is fairly air tight. If your heat load is 45K @ +70F in/+20F out (a 50F difference), your heat load wouldn't reach 140K until it's ~100F colder than +20F, or about -80F outside, a temperature not seen in Raleigh since well before the last ice age. (That's roughly the average annual temperature at the south pole, which is also 9,300' above sea level, and far from the moderating temperature effects of oceans.) Seriously, run the fuel use base heat load calculations- whatever number that ends up being is a firm upper bound- some of that fuel heat went into your hot water, then down the drain. Your real heat load will be somewhat lower than implied by the fuel use.

Flex duct is cheap & easy to install, but fraught with installer error issues, and have much higher back pressure issues than metal duct, especially when mis-installed. Any seams or joints that can be reached can be sealed with duct mastic (foil tapes only seal well if the metal is very clean, which it surely isn't after 34 years of service.) Unless the duct "sweats" during the cooling season (and thus needs to be replaced for moisture management) the energy savings and peak load effects of going from R3-R4 rigid duct to R8 flex will be pretty tiny, and the costs of the new ducts will never be recovered. Sealing any accessible duct seams with duct mastic is usually "worth it", if duct-blaster testing detects high leakage rates. Air sealing the duct boots to the ceiling gypsum would worthwhile/necessary even with new ducts.

 

[Themp]

Dana, uploaded an image of my gas bill history. Does this have the right data?

https://www.dropbox.com/s/ryv43zkyqjlkh8e/File Aug 08, 1 14 47 PM.png?dl=0

 

[Dana]

If I'm reading it correctly...

12/23/2015 & 1/26/2016 -----------119 therms

1/26/2016 & 2/23/2016 ------------154 therms

Total for the heaviest 2-month period is 273 therms, or (x 100K=) 27, 300,000 BTU. Burned in an 80% efficiency furnace that delivered 21,840,000 BTUs into the ducts (the rest went up the flue).

Since you didn't share the ZIP code I'll use data from weather station KRDU (the Raleigh Durham airport) on degreedays.net.

From the interval of 24 December- 23 February they logged 1336.1 heating degree days

So, your house uses (21,840,000/1336.1=)16, 346 BTU per degree day, or (/24=) 681 BTU per degree-hour.

The 99% outside design temp is 20F, the presumptive heating/cooling balance point is +65F, for (65F -25F =) 45F heating degrees.

The implied heat load is then 45F x 681 BTU per degree-hour= 30, 645 BTU/hr.

That's about right for a fairly tight 1500' + 1500' 2 story house with a fairly simple footprint such as a rectangle, not too many ells adding corners and exterior surface area. I would also expect that there isn't a lot of exposed above-grade foundation for the basement area for it to come in that low.

If you normally keep it 65F most of the time, use base-60F, and add another 3F to cover for the difference between 65F and the code-min 68F heating interior temp. Using base 60F they logged 1066.4 HDD, for

21,840,000/ 1066.4= 20,480 BTU/HDD or 20,480/24= 853 BTU per degree-hour.

The difference between base 60F and the 99% outside design temp of +20F is now 40F, but to meet code-min the heating system would have to be able to heat to 68F, so make that 43F heating degrees.

43F x 853 BTU per degree-hour= 36,697 BTU/hr

That's also a credible number for a house that size.

Unless you spent half of January on the beach in Zanzibar with the T-stat turned down to 50F or something, there's no way the design heat load is more than 40,000 BTU/hr. It's almost certainly going to come in between 30.7 -36.7K BTU/hr if you did an aggressive Manual-J.

You can fine tune it base on weather data from a station closer to your house, but that's unlikely to skew it more than 10%.

So, if you were going by the ASHRAE recommended 1.4x oversizing factor the optimal heating equipment would only need to deliver deliver 43-51,000 BTU/hr. A single condensing furnace with 55,000-input would cover the entire heat load right on down to 0F without losing ground.

Yet your contractor was recommending a 60K for one floor PLUS another 80K unit for the other, more than 2x the capacity you would need even during the coldest hour of the past century, which is just plain ridiculous.

Depending on vendor & model, 2.5-3 tons of air source heat pump could cover a 35K heat load at 20F, with some resistance heating strip backup to cover for the Polar Vortex extreme events when the heat pump alone wouldn't quite make it.

 

[Themp]

Dana, thanks for the very detailed analysis. You said:

A single condensing furnace with 55,000-input would cover the entire heat load right on down to 0F without losing ground.

So, does this mean one 55K BTU furnace could handle the complete house? If so then if I did not do the heat pump solution, an 80% single stage furnace for both floors could be in the 35K BTU range. Two stage furnaces could go lower than 35K BTU running on the lower stage.

I did see some discussion of a dual fuel heat that uses a heat pump and then has gas backup heat instead of the electric strips.

 

[Dana]

Yes, and single 55K-input/50K-ou unit could more than handle the complete house, even at the coldest temperature of the past 25 years. A pair of 35K units would be sub-optimally oversized, but way better than what you have.

Most simple 2-stagers at the low firing rate only drop back to about 60% of max. So, 0.6 x 35K is 21K, and 2 x 21K = 42K, (x 80% efficiency= ) 33.6K which is approximately 99% temp load. It would only need to kick to high-fire less than 87 hours per year to keep up. After you fix up the ducts and air seal the attic you may literally NEVER need them to kick up to high fire at that oversizing factor. But it's hard to find 2-stagers much smaller than that.

If you're dropping back from 100K to 35K of burner per furnace the existing flue is almost certainly oversizes, and it would need a narrowing flue liner to draft properly and avoid excessive flue condensation. If the flue has to be retrofitted to be able to use 80% efficiency equipment it can sometimes be cheaper to go with a condensing furnace and plastic vent pipe.

Dual-fuel gas + heat pump are more commonly used in colder places, such as northern New England, or the Upper Midwest, where 99% outside design temps are well into negative double-digits. At those outdoor temperatures most air source heat pumps have much lower capacity, or can even self destruct, if not designed for sub-zero operation. Dual-fuel units are designed to be set up switch over from all heat pump to all gas burner at the outdoor temp at which burning gas is estimated to be cheaper than running the heat pump. The optimal crossover point will change with every change in gas or electricity rates. Heat pumps with electric backup usually still run the heat pump while the heat strips are running, and most of the BTU output is still being sourced by the heat pump, even when it's super-cold out.

 

[Themp]

Dana, again thanks for all the analysis. I am starting over looking for contractors at this point. Your flue comments are great also, would not have known this. Printed this thread off and it will interesting to see what the contractors have to say.