GeoMax 2 (rebranded CM) 4-ton with two 300-foot boreholes, Rehau PEXa in double-U-bend configuration (4 loops total).  Various problems in the past detailed in various posts (duct work, faulty wiring, wrong anti-freeze mix, etc.).  Propylene Glycol antifreeze, 2000 sq foot home, 2 zones.

Finally got WEL installed and beginning to add sensors.
EAT and LAT are in the main trunks right by the unit.
EWT and LWT are attached to the brass distribution manifold - no insulation around them yet, so may be affected by basement temps (around 58F ambient).

Plan to add outdoor and indoor sensors, possibly zone 1 & 2 separate, and either a run-monitor current switch, or tap the accessory relay on the CXM board for unit on/off status.

Advice, tips, tricks, welcomed, as well as serious perusal of initial data. 
Temperatures of the WEL sensors seem "off" by a few degrees that other thermometers I was using, but I'm not sure which ones are accurate.  May have to calibrate WEL sensors with ice water/etc.

My WEL page  WEL0487

Stuart

Congrats.

You could meassure the water temps right through the p/t ports, and then use that to calibrate your sensors for the EWT and LWT. That way the ambient air is totally out of the equation.

Stuart, I recommend:

1.) You shorten your charts horizontally to about 1000 pixels.  This will eliminate the need for a viewer to have to scroll left and right to see the whole chart.  This would also put the Y axis on the screen too.

2.) Use the Sample and Hold function so that your numbers are representative only of 'on' time.

Nice job so far!

Best regards,

Bill

It would be interesting to collect enough data for real-time COP.

a012... with the exception of estimated extracted kbtu, all the numbers should be doing the sample & hold thing. I'm using the zone calls as the "trigger" to sample vs hold. What I think is happening though is that the system goes on, but the sensors take a minute to get to the actual loop/air/etc. temperatures, causing a weird spike. I almost need to delay the sample-taking by a minute.

I'd love to hear if other people have had success doing it a different way. Mine is just: EWT_raw "S" SystemOn = EWT_display (WEL programmers, that should make sense to you :-)

Sorry about the width...widescreen here! Will adjust.

Posted By jonr on 28 Apr 2011 10:38 PM
It would be interesting to collect enough data for real-time COP.

I do.  This is a good example of the advanced computational capabilities of the WEL (i.e. it's not just a real time value display system, or just a data logger).

When my WaterFurnace Envision units (I have two - a 5 ton and a 3 ton) are in heat mode, my WEL unit calculates COP, displays it on my real time image ( http://www.welserver.com/WEL0043/ , it's also in my signature below), and logs the data every minute for later analysis.

For the month of Jan. (my coldest month this year), my 3 ton unit ran at about 5.25 COP.  Using logged data from the WEL, and doing a histogram analysis on it, here's what it looks like.  Keep in mind this is first stage only - I don't know what 2nd stage looks like because I so seldom run in 2nd stage.


For the same month, my 5 ton unit ran at about 4.75 - 5.0.  Again, this is first stage only.



My WaterFurnace Envision units were installed in 3Q07, so they're about 3.5 years old.  I have an 8 ton max capacity (for my Dallas location) borefield so my WaterFurnace units are not limited in performance by the borefield's capacity.  What you see here for COP analysis is what the WaterFurnace equipment itself is capable of producing.

Hope this helps.

Best regards,

Bill

to do COP, I would need to measure electric usage, correct? I can't, for example, divide delta air by delta water and get a COP?

Posted By decafdrinker on 29 Apr 2011 07:42 AM
a012... with the exception of estimated extracted kbtu, all the numbers should be doing the sample & hold thing. I'm using the zone calls as the "trigger" to sample vs hold. What I think is happening though is that the system goes on, but the sensors take a minute to get to the actual loop/air/etc. temperatures, causing a weird spike. I almost need to delay the sample-taking by a minute.

I'd love to hear if other people have had success doing it a different way. Mine is just: EWT_raw "S" SystemOn = EWT_display (WEL programmers, that should make sense to you :-)

OK.  Looking more closely, now I see that you're using sample and held values.

Next step to consider for reducing the noise spikes is filtering.  I'd put a 2 hour filter on EWT and LWT related numbers.  The 2 hour filter is a WEL built-in function.  If you want to go shorter, i.e., using a 1 hour time constant, you can do some WEL 'programming' to accomplish this.

For your EWT and LWT charts, you actually may want to consider using a 1 day filter.  I.e., show the 24 hour moving average of EWT and LWT (this is what I do - see http://www.welserver.com/cgi-bin/pl...ilyAvg.gif ).  You can use your real time display to see minute changes in EWT and LWT if you want to, and chart what seems to me be make sense - a 24 hour moving average.

Yes, the WEL has a 60 second delay between run-time switch inputs, and sensor value collection, such that sample and held values will have some noise in them.  Just filter them.

Hope this helps.

Best regards,

Bill

Posted By decafdrinker on 29 Apr 2011 03:38 PM
to do COP, I would need to measure electric usage, correct? I can't, for example, divide delta air by delta water and get a COP?

Yes, for a given unit, you need to measure electric usages, water flow rate, and EWT/LWT DeltaT.

Best regards,

Bill

We use the current sensor to measure amps drawn, we calibrate it at stat up.
http://www.greenenergyresearch.com/DIY_ph_pages/DIY_ph_3502.html?product_id=3502

It is quite accurate, we run it in parallel with the wattsnode, and it is within 0.5% after calibration. Here is one with both the current sensor and the wattsnode. http://welserver.com/WEL0396/

So water flow rate is constant, so is voltage, the current sensor measures your amps, and your temp sensors give you your delta T for the source. The rest are formulas, and you got your COP.

I'm sure it works well, but if for some odd reason you want really close accuracy you need to account for voltage changes, flow changes (viscosity changes with temp), specific heat and power factor. Then there is the heat that comes off of the motor and compressor - some goes into the air, some goes into the system and gets measured.

$35 for the current sensor vs $240 for the wattsnode, $550 for a basic welserver kit, $8000 for a monitoring system -$14000 for a monitoring system which can do everything in real time as you suggested. I thought I mention my experience. BTW, if you have a rounding erro of 0.1 degrees F at 5 degrees F delta T for the source, your numbers will be off by 2%, 4x as much as with the current sensor. And I still have to get someone to give me a really good number for the power factor. How much is voltage really gonna change? I really don't have much intention to re-calibrate everything daily or weekly. If I am within a couple % of real numbers, that is good enough for me. The rest I leave up to engineers!

I really appreciate all the wisdom and knowledge about COP and the various ways of calculating it.  I can't help but wonder if someone has the most accurate measurements and they now know that their COP is lower by 2-% -5% than they thought it should be, what would they do next?  What is the cost of high accuracy and the return on that investment?  There is always a ROI to consider.   FIrst cost accuracy is more important to me than whether we got a customer 5.1 COP or 5.15 COP.  If that .05 pays back the added first cost to have COP accuracy withing 10 years, I guess accuracy might be worth considering.   Happy Monitoring!

The WEL server has allow us to significantly increase the efficiency of our systems, and to get feedback on our systems installed. What is the EWT at the end of the heating season with a certain kind of loop? We were able to test different loop designs and significantly reduce our installment costs. We get feedback on the run time, loop performance, heat extraction, DSH performance, circulation pumping power, impact of flow on performance, Amperage drawn under certain operating condition (temperatures), especially outdoor resets, etc. All this has allowed us to increase operating efficiency while reducing our installment costs. It sounds strange, but I am really not interested in the COP, but instead in all the parameters influencing the COP. As an example, if I can reduce the supply temperatures to a radiant system by bringing the radiant pipes closer together, by how much does that allow me to reduce the leaving load Temp (which increases the COP). As a result we now design some systems with a maximum LLT of 85 F.
Or the effect of certain backfilling techniques on loop performance in the first heating season after installation (we are oversizing our loops now, which turns out to reduce installment time).
Or the stacking of circulation pumps (4 tons and up), having 1st stage come on with 1 pump, and the second pump come on with the second stage. It takes one relais and a bit more wiring, but it safes over $100 per year in operating costs.
So we play with things, and then wonder what impact this has on performance. It also has helped us to fine tune systems (for example outdoor reset curves).
Overall COP is a result of those measures we monitor, but people are too fixated on COP, and don't pay enough attention and try to understand the things leading to it.

Posted By docjenser on 01 May 2011 01:21 AM

....... (we are oversizing our loops now, which turns out to reduce installment time).

Hey Doc,  this interests me.  Would you please elaborate?

if someone has the most accurate measurements and they now know that their COP is lower by 2-% -5% than they thought it should be, what would they do next?

Decide that is it time to clean their heat exchanger (in an open loop system)? Activate a soaker hose for an hour? Check refrigerant levels? Change a design for future systems? Lots of things, especially for someone involved in tweaking designs (an excellent idea - see above).

But no question that a negative ROI would fall into the category of "odd reason" (see above) for a typical homeowner.

Posted By gonegeo on 30 Apr 2011 10:25 PM
I really appreciate all the wisdom and knowledge about COP and the various ways of calculating it.  I can't help but wonder if someone has the most accurate measurements and they now know that their COP is lower by 2-% -5% than they thought it should be, what would they do next? 

As Doc said better to understand the contributors to COP than obsess over a specific number.
Most folks would see COPs that they thought were low if they measured in real time all winter.
There is a reason average COP's are 3.5 to 4 in most systems (heating dominated climates)....it is because sometimes in deep winter your EWT might be at it's annual low and you might see significant aux. use.

Further .5 or 1 COP can mean little cost or savings depending on the price kwh. Doc and I had a lively discussion about his oversized loops which make sense in his context (high elec cost, different type of soil/compaction requirements) but not mine (chep elec and easy compaction). 

Design practices also vary with electricity cost, so COP is likely to be lower with my design (cheap electricity allows more auxiliary contribution) than say a long island installer who is working with kwh that cost 3 times as much.

So while awareness of a systems performance is great, automatically assuming trouble or taking corrective action if COP is "lower than expected" is folly. Sometimes nothing at all is wrong.

Joe

Posted By geodean on 01 May 2011 09:58 AM

Posted By docjenser on 01 May 2011 01:21 AM

....... (we are oversizing our loops now, which turns out to reduce installment time).

Hey Doc,  this interests me.  Would you please elaborate?

Dewayne, http://www.greenbuildingtalk.com/Forums/tabid/53/aff/13/aft/77624/afv/topic/afpg/1/Default.aspx

Biggest impact on COP is to change the operating conditions of the heatpump, lowering the load temperature with W-W units. On the whole system, I found that circulation pumps are energy hogs. Not many people monitor their systems, I think it should be standard, and the way manufactures are talking, it will be incorporated on the circuit boards soon. The systems we see nowadays have a large forgiveness and a large safety margin. I know, hard to imagine that we see a good number of people here having problems. It is unimaginable to me that I would leave a customer with a poorly performing system.
Anyway, most contractors leave and have no idea if the system is performing well, how good their heatloss was, how much aux heat kicks in, how much the actual run time is for every stage, did I build the system too large or too small. So they have no feedback on their jobs. Their outcome measure is whether the customers call back and complain. Not really a precise way. So a monitoring system like the WEL for $600 working within a couple % of real numbers should not trigger any accuracy concerns, but instead should be considered a great asset.

Absolute accuracy is less important than precision - in other words, the ability to perceive and react to changes in system performance - running a soaker for an hour during a drought or flushing a coil with Nickel Safe upon seeing a dip in performance.

Reading Doc's bit about dumping an unneeded circ pump rings true. I've taken a similar smaller step with a zoned system - zoning supposedly always requires an extra 24 Vac control power transformer to provide extra power for board and dampers. I measured VA on mine at times of max damper power use and found it well below VA rating of a single control power transformer. Out came the 2nd transformer which was using about 10 Watts at idle per TED.

10 Watts isn't much, but control power transformers are a 24 / 7 / 365 load. Losing 10 Watts is worth $10-20 per year, depending on kWh rate. Know that some dampers use more power than others, and some use power to hold a closed position, so I don't recommend automatically doing as I have without data.