Joe,

I think you are on to something with hypothesizing a relationship between setback payback and loop length. A marginal loop will be at its absolute worst during the long continuous run needed during recovery. Savings from lowered building load could be overwhelmed by COP depression during a long run.

I think it worth observing that heat loss isn't a function of thermostat setpoint, but actual building temperature. If a building at 70 is setback to 65 with an outdoor temp at 20, its loss will go down 10%, but only after the interior temp completes it drop to 65. In the case of large, modern houses, that might not happen until hours into the setback.

If such a house lost 1 degree per hour for 5 hours, its average temperature during that time would be 67.5, corresponding to a building load reduction of just 5% for the entire period.

Of course, some loss reduction savings continues until recovery is complete.

If one accepts the relatively low savings rate in the above example as reasonable, the next thing to consider is whether otherwise avoidable stage 2 operation occurs during recovery. Stage 2 operation carries with it lower COP, both owing to depressed loop temps and machine efficiency loss (higher blower watts per CFM, higher refrigerant flows and delta-Ps). If stage 2 COP is 20% lower than stage 1, then it won't take long to counter a 5% load savings if skipping the setback would have reduced / prevented stage 2 operation.

Running the above numbers for a milder outdoor temp of 45 doubles the savings (each degree of indoor temp drop reduces building load by 4% instead of 2%), and may reduce stage 2's offsetting loss of efficiency.

1)   The day is question,  was a bright sunny day,  and I have  a lot of south facing windows.

2)  My house is a simple floor plan,  pretty much a box, and I being the builder, I took extra care to make it tight.

Dewayne, thanks.  As I suspected.

I created a chart exactly like one of yours: http://welserver.com/cgi-bin/plot/WEL0043/LoopTemps24Hrs.gif , showing room ambient temperature along with outside air temp and GSHP cycles.  For a similar outside air profile but a little warmer, I run through about 40 cycles for a complete day, compared to your handful at most non-setback time period cycles ( http://welserver.com/cgi-bin/plot/WEL0114/LoopTemps24Hours.gif ).

I'm almost complete with a number of days running a constant 68° F tstat setting.  Since repair of my GSHP unit doesn't appear to be scheduled for the next few days, I may switch back to my standard-in-the-past 3° night time tstat setback-recover scheme.  I'm averaging pretty consistently a 1.5 KWH/HDD index score for the past few days with my constant 68° tstat setting, so I think I have a pretty good baseline to compare to.

Best regards,

Bill

Joe,  I appreciate your input.  I welcome all questions and constructive feed back.

My house was my first geothermal project.  The design was done by others.

1) There is no question that my geosystem is oversized,  both the ground loop and heat pump.  My loop temp has never dropped below 40°  or risen above 70° through 4 years of service.   It wasn't until I installed my WEL monitor that I realized how over sized my system is.  I have learned a lot about system and loop design through the discussions on this forum. (I used to think that using aux heat was an indication of poor design)

  On systems that I design and install,  we will be sizing the loop and heat pump closer to design temps in an effort to  lower the install cost and broaden the appeal.

I am not sure of the impact of longer loop on set back recovery.

2) Not sure what you are asking here.  Are you referring to loop temps?   If the question is,  does it make sense to spend extra on the loop to save some on operating costs,  I am not sure.

We are in a unique position in our area.  There is hardly any competition.  Since geo is in it's infancy here and I am still learning the finer points of system design,  I am probably guilty of erring on the side of caution when sizing pumps and loops.  I have been driven by the intense desire to not install a system that would not keep the house warm on a cold day.  My mantra has been,  "one unhappy customer will set back this emerging industry for years."


I am going to modify my approach now ( thanks to you guys )  and install leaner systems which might use aux heat once in a awile.

3)I don't think that we can extrapolate my results onto any other system.  I am anxious to here what others have to say.

4) If a system is frosting,  I doubt that set back  will improve the situtation.  Might even aggravate it for the short tem with longer run times.

5) I think the best way to approach this is to give the customer a choice when bidding.

     One system that covers  85% of the load with 15% aux heat  for $xxxx

     One system that covers 95% of the load with  5% aux heat  for $yyyy

     One system that covers 100% of the load with 0% aux heat for $zzzz

On another board there have been geo customers asking why they were not given this choice when bids were presented.  I kind of like the idea.

**********************************

I want to say how much I enjoy the discussions we have on this board and appreciate the knowledge and experinence that is shared.

One other factor that I forgot to list:

My tstat is set with a 3° differential.  I did this to minimize run times.  So in the morning,  the heat pump runs till the house is 71° then won't come on again until the temp drops to 68°.  If the solar gain keeps us above 68°,  then the heat pump won't get called for heat.

If my  heat pump were smaller, it might run  most of the day to recover from a 5° setback,  if the day was cloudy and there was no solar gain.

So more results to digest.

First day of test    HDD = 31     KWH / HDD = .66

Second day          HDD = 43      KWH / HDD = .75    

Third day             HDD = 47      KWH / HDD = .62

Fourth day            HDD= 38       KWH / HDD = .98


The fourth day was cloudy and windy with no solar gain.   My  heat pump ran 5 times on day 4 compared with 2 times on day 3, even though day 3 had more HDD than day 4.

Sun and wind  are really messing with our attempt to quantify the effect on cost of set backs.

Awaiting your analysis...

Joe,

have you seen the charts here?

Dewayne,

I really like the idea of giving the customer a choice for heat pump versus auxiliary heat.  My experience has been that the customer did not know he had a choice.

The solar gain is what jumped out at me. Until you have a bunch of cloudy days in there (maybe there are?) wouldn't it be hard to know how much extra heating your getting from the sun? At the same time if you could compare similar (same HDD and same wind) sunny days with cloudy ones you could extrapolate BTU from solar gain correct?

A bit off topic but do HDD's take in account wind? I have noticed when we have wind out of the Northwest (the only no wind protected side of the hose) I lose temps faster than any other time. It seems at 15F wind a 15 mph wind out of the northwest wind will require more BTU's then 0F and calm. Although I suppose that would vary wildly depending on construction.

I've ended up getting a pretty good baseline for holding a constant 68° F throughout my residence.  For the past 4 days, I've consumed 144 KWH for heating, and have had 102 Heating degree-days.  That's an index average of 1.4.

For each day, the index values have been:

Sun     22.7 KWH     17.5 HDD     1.3 Index     48° avg outside temp

Mon     31.9 KWH     22.8 HDD     1.4 Index     42° avg outside temp

Tue     44.6 KWH     29.4 HDD     1.5 Index     36° avg outside temp

Wed     44.4 KWH     32.5 HDD     1.4 Index     33° avg outside temp

Some days have been bright sunshine (yesterday), some days have been no sunshine and a lot of rain (Tue.), some have been cold (yesterday) and some have been warm (Sun.).  Overall, my daily index values are quite close to the overall average, so I think I've got a pretty good baseline to compare tstat setback-recovery to.

Today will be the last day of the tstats at a constant 68°.  This will complete a 5 day baseline.

Tonight at midnight I'll flip the tstats to setback-recovery mode, and then I'll use tomorrow (Fri.) as a day for the structure to acclimate.  Sat. will be the first day to collect data for a setback-recovery tstat setting.  This data will be available on Sun.

I'll be returning to my 3° setback-recovery tstat scheme.

My accumulated index value for the day, KWH consumption, and HDD are all on the first page of here: http://welserver.com/WEL0043 .  And the last 2 charts show tstat temps.

Best regards,

Bill

Dewayne,
apologies, haven't seen charts anywhere. It's a cross between living in a dial up area and mostly using my wifes computer which would get my butt whipped for down loading anything. I did like your question about offering different capacity systems but want to emphasize the point that one of the benefits of a little less geo and more aux. heat fits duct systems better.
There is almost no new construction here right now so most of our work is retro. That said we have to deal with duct system we have or charge geometrically more for duct modification as well as extra ton(s) of geo.
It is usually no contest.
Appreciate your tests, if you don't mind typing a "Cliff's note" I'll know more than if you refer me to a site.
J

I completed a 5 day baseline of maintaining a constant 68° F for my residence 7x24.  Key data is as follows:

Sun     22.7 KWH     17.5 HDD     1.3 Index     48° avg outside temp     173 KBTU consumed

Mon     31.9 KWH     22.8 HDD     1.4 Index     42° avg outside temp     263 KBTU consumed

Tue     44.6 KWH     29.4 HDD     1.5 Index     36° avg outside temp     373 KBTU consumed

Wed     44.4 KWH     32.5 HDD     1.4 Index     33° avg outside temp     376 KBTU consumed

Thu     37.9 KWH     25.9 HDD     1.5 Index     39° avg outside temp     311 KBTU consumed


Total KWH consumed was 181 for the 5 day period of 128 HDD, producing an avg index of 1.4 KWH/HDD.  Each day's individual index was within 10% of the 5 day avg index.


Today is a transition day to let the structure adjust to the tstat's setback-recovery scheme.  Tomorrow will be the first day to collect data.

To not overly complicate the experiment, I've chosen a simple setback-recovery scheme: maintain the entire structure at 68° during the day, and set back 3° at night.  The residence will shift from a 68° to a 65° setpoint at 10 PM.  All living areas will then shift back up to 68° at 6 AM, Bedroom 1 will shift at 5:30 AM, BR2 will shift at 6:30 AM, and BR3 will shift at 7:30 AM.


Parenthetical notes:

1.) My tstats are set to a 1.2° heating differential (largest I can set tstats to), and set to not allow 2nd stage to turn on until 40 min. has elapsed in 1st stage.

2.) My structure is 3400 s.f., all single story.

3.) Window treatments are being opened/closed at same general time points each day.

4.) I don't have any aux heat capability.


Best regards,

Bill

Dewayne ~ as a homeowner I really appreciate your idea of giving the % aux options. When I was looking for estimates, I was told that I needed a units ranging from 3 tons to 6 tons ~ it made for a confusing selection process, but also forced me to learn more before I made the decision.

Bill ~ thanks for running this experiment!

Bill,  congratulations on gathering some meaningful data.  I have not very good luck, due to the high solar gains in my house.

It looks like we will get some meaningful data from your efforts

I am wondering since the solar gains at my house have made my study somewhat meaningless,  if I should try another approach.

If my tstat is set back  5 ° from 9 pm to 4 am ,  then the heat pump has to make up the  heat lost during the set back the first time that it runs during the day consuming xx kwh/hdd.  This is usually finished between  7 and 9 am.


If my tstat is not set back, then  the heat pump cycles on and off during the night  keeping the house at the set point and consumes yy kwh/hdd.

Do you think that we can compare   the xx kwh/hdd with the  yy kwh/hdd and get any meaningful data?

The hard part will be accumulating the data over two days,  when the WEL reports data daily.

What about changing the time on the WEL so that the data starts accumulating at 9 pm instead of midnight?

Just thinking out loud here,  any and all comments will be appreciated.

Posted By geodean on 01/30/2009 11:31 PM

... I have not very good luck, due to the high solar gains in my house.

I've been thinking about this.  I think there are two problems:

1.) The experiment should be run with window shades drawn for the full duration, to maximize the BTUs needed to heat the structure to be coming from the GSHP.

I didn't think to do this.  I believe that variation in my data is due too in part to solar gain.  But, the variation is minimal because of the second problem:

2.) The experiment is dependent on how 'normal' the structure is with respect to air tightness.  My residence is 'leaky.'  And thus the contribution to BTUs needed to heat the home, from solar gain, is not as significant for me.

Dewayne, in your case, it can be 20° something outside and yet your ambient temperature still rises with solar gain from a very bright day.  There's no way my home could ever accomplish this.  So 'air tightness' appears to be a significant influencing factor too.

Tomorrow will be my first day to collect data for a set back scheme. 

Best regards,

Bill

I think this is pretty insightful.  Basically I think you're saying, Dewayne, is, let's collect data only when the sun is not shining.  And the question basically is, can this be done with the WEL?

The hard part isn't hard.  The WEL reports data on a 60 second basis.  It's all in the log.  xx and yy are both defined by a starting and ending time.  Note the KWH_D starting point and substract from the day's total, reported at Midnight.  Then note the KWH_D stopping point, and substract from zero (beginning of new day).  Add the 2 numbers and you have xx.  Same for yy.

Don't know of any way to change the WEL's accumulation start time to anything but Midnight.  But by using the minute-by-minute log, you can get the interval of anything defined from a finish and a start, including going across the Midnight boundary.

In fact, you could restate your already published number, as night-time only numbers, by getting the already existing data from the log, and then seeing if the numbers have better consistency.

Best regards,

Bill

On the first page of the WEL you can set the time,  see below.

I am thinking that if I set the time to 9pm when it is really midnight,  then the WEL will start the new day at 9pm instead of midnight.

Ok ....I changed the time on my WEL so that it thinks it is midnight, when it is really 9 pm. I will collect KW/HDD for the first 9 hours of the day using a set back for 5 days and then no set back for 5 days.

I hope this will eliminate the interference from solar gain.