Okanogan 14x24 by a lurker :)

[OlJarhead]

OK, starting to learn more and more but still much to learn. 

Arrived at the cabin just before 6pm last night (long day at work) and the generator was running.  Not having any idea how long it had been running I set the dial to 5hrs because I knew I'd want to abuse the power with it on (cabin runs off the genset when it's running) and figured it wouldn't hurt to run an extra hour or two to float the batteries a while.

I didn't really abuse the power much but left the gen running until about 10 PM (maybe 9:30 - should have written it down) and went to bed later having checked the voltage but not recording it (I swear I did but can't find it now).

This morning the voltage read 23.9v at 7am and I did some calcs based on previous findings as well as some checks last night and figured I'd used 35AH overnight.  I then checked the meter and it read 92% of cap based on 420AH I set it at.  I then did some quick math an came up with this:  420AH - 35AH = 385AH and 385AH = ~92% of 420AH

So the meter reads the amp draw over time and shows percent of capacity based on the setting you gave and the amount drawn from the system over time.  Makes sense to me now actually and should be a good indicator of actual capacity under use than voltage or PG readings are since both are inaccurate unless the bank has been sitting completely idle for at least 6 hours (preferably 24hrs according to Trojan and Trimetric both).

Now, had the batteries been sitting idle for at least 6 hours and remained at the 23.9v then the bank would be 60% discharged not the 8% calculated.

This appears to be where the biggest confusion is as to where to set ones generator to run and has been the center of many discussions on these boards for some time.  If you set the generator to come on at 24v which I originally did because I assumed that at 24v I would be about 50% DOD (depth of discharge) which I didn't want to go beyond and because I could not set the switch to start higher, than according to Magnum Research, AIMS Manufacturing, TriMetric and even Trojan you're setting the switch too high for a system under load and overusing your generator.

Magnum and AIMS both said something on the order of 22.5 would be quite safe but I honesty couldn't wrap my head around it because I've never draw the bank down to that voltage and then stopped charging/discharging and let it sit long enough to check voltages and SG readings to see what happens.  By extrapolation it sounds like what should happen (should being the operative word) is that the batteries should recover a fair amount of voltage and sit much closer to 50% DOD instead of the 100% you'd think.  Note:  23.02v on a Trojan T105 is 90% DOD for open circuit voltage and after a rest.  Their note reads:  For accurate voltage readings, batteries must remain idle (no charging - no discharging) for at least 6 hrs, preferably 24hrs.

I think this is the key to much confusion.

So at this point I at least understand how the meter works and can understand what I'm reading now and can at least adjust the meter to tell me the percentage of cap remaining based on a battery bank at the lowest temperature I'm likely to see while here.  Meaning that even though I should have 660AH of capacity if the bank drops to 32 degrees Trojan tells me their GCB's are at 65% of cap which in this case would be 420AH.  So by doing this I can set my generator to come on at 85% of cap (it's a voltage setting but I can watch the cap % drop until it hits that 85%) of a battery bank my size at 32F and I should (should again being operative here) be safe and not abusing the bank too much.

Of course I have to consider that the bank is likely sulfated a fair amount and even with 'Charge It' in the system it's undoubtedly at lower cap than I think unless a good EQ charge and the Charge It works to get rid of most of the crystals.

Today I plan to disconnect the battery bank once it's charged up and I'm ready to do something else, and let it sit for 6hrs, then I'll check the SG readings and voltages for each battery (and record them) and see where it's at and if it's really out of whack.  Next I'll equalize the system and see what that does to overall performance.

What am I looking for here?  I'm looking for good and even SG readings and a bank that shows 25.46v after a 6hr rest an having been fully charged.

Somehow I doubt I'll get it but then who knows?

[OlJarhead]

OK, messing with the Tr-Metric meter I set it a little closer to what I think it should be.

1.  Charged setpoint voltage:  29.2v (Based on the absorption charge target of 29.6v from Trojan's T105 Battery -- no I don't have T105's but they are the closest I could find to my Costco GCB's)
2.  Charged setpoint Amps:  10Amps (this one is tricky and is set at 2% of battery bank AH cap which I changed to 490AH based on 40-45F which the bank sits at most of the time)
3.  Battery Capacity in AH:  490 (Yes I changed from the 32F above but after thinking about it realized they are staying at 40+ most of the time now and will likely.  Of course, this is not the 660AH they are supposed to be so I might be doing this wrong but we'll see).
4.  Primary Display:  Amps (vs Watts)
5.  Days before 'time to recharge':  set to the recommended 5
6.  Days to Equalize:  OFF
7.  Operation Level:  L1 (this one I need to investigate next).

[OlJarhead]

According tot he historical data on my Tri-Metric meter (which I've only just checked for the 1st time) it has been 5 days since the system was last fully charged (assuming I am reading it right) or the cycle length (from charge 100% to discharge to charge 100%) was 129hrs.  However this is odd since my MPPT controller tells me that 5 days ago the max voltage was 30 volts and the system went into absorb for 1 hr but did not float.  This should in theory have been fine since the system was not in use.  Battery low voltage went to 24.42vdc with only a very small draw on the system (DC fans running in back room that draw very little -- I need to check but I'm thinking something like 10ths of an amp) so the system should have easily hit 100% of charge/cap every day for the last 5 days.  Here are the AH's produced by the panels over this time:

Amp Hours: 16.8, 24.2, 21.8, 26.2, 26.7

So each day during this period the PV system produced plenty of charging to recharge a bank that is in very limitted use (almost not used at all since only the Tr-Metric meter, MPPT Controller, 24v-12v converter and two 90mm 12vDC fans are running when I am gone --I'll find what that draw is shortly as I Can do this by merely killing the Inverter and MPPT controller to see what the draw is).

The Minimum voltage on the meter shows 20.9 volts (which is a concern) but the MPPT controller shows the min volts during this period as 22.46v.  Now the controller does compensate for temp but I don't think that should change the voltage reading -- am I wrong here?  There is a 1.5v difference in the two which concerns me.

I'll go look to see what it is now and report back...but before I do I observed some other things:

The length of cycles in hours shows: 129 (most recent), 556, 47.2, 72.9 and 86.5 (most distant)
This suggests that the system say 100% of cap/charge 129 hours before I took the readings which I find strange since I ran the generator on the Thursday I arrived which is just 41 hours (roughly) since since charging the system to 100% on Thursday but the system reached 100% twice on Friday and again this morning so if I understand the system correctly it should have read only about 10-12 hrs for the last cycle.  Perhaps I'll reset the system and see if it shows me tomorrow that it's been less time.

[OlJarhead]

A quick check of voltages reveals:

27.09v MPPT Controller
27.1v   Tri-Metric Meter

So this appears to be the same (virtually)

3.7Amps Charging on the MPPT Controller
2.9Amps Charging on the Tri-Metric Meter

Currently the generator is running as well as the MPPT Controller but the gen tends to drop to 0 when the system is fully floated and the Controller is charging in float.

So the voltages reported are basically the same at this time so I'm confused why they would be different over the last 5+ days.  I'm going to reset the system to see if I can start fresh while I'm here.

[OlJarhead]

OK, so after charging the system to what should be a full charge I took SG readings to see how bad things were.

1st string: 
Batt 1:  1.225/1.225/1.225 @~45F so possibly 1.209-1.211
Batt2:   1.225/1.225.1.225
Batt3:   1.225/1.235/1.230
Batt4:   1.220/1.220/1.200

2nd String:
Batt 1:   1.250/1.250/1.250
Batt 2:   1.240/1.240/1.240
Batt 3:   1.200/1.200/1.200 (at this point I changed SG testers because I was having trouble reading the tester accurately.
Batt 4:   1.220/1.220/1.220

3rd String:
Batt 1:   1.235/1.240/1.235
Batt 2:   1.220/1.220/1.220
Batt 3:   1.222/1.221/1.230
Batt 4:   1.221/1.221/1.230

At this time I realized those readings were pretty much all over the place and decided to equalize so I set my inverter/charger to Equalize (desulfator selection) and let it run for 3 hrs (the default time for an equalize on my MPPT controller.  The inverter times out at 4 hours but I decided to go with 3 hrs to see how it turned out.

Afterwards I took some more readings:
String 1:
Batt 1:   1.230/1.230/1.232  (rise of 0.005+ on each cell)
Batt 2:   1.226/1.228/1.230 (rise of 0.001 on cell one, 0.003 on 2 and 0.005 on 3)hmmmmm
Batt 3:   1.234/1.234/1.236 (rise of 0.009/-0.001/+0.001)
I was then interupted for a good 30 mins and the voltage on the bank had dropped before I could get back to the bank and so my next reading may or may not be accurate:
Batt 4:   1.218/1.222/1.218 (-0.002/+0.002/+0.018)

I decided at this point that it might not be the best indicator since time and system usage had taken place from the 3rd to the 4th battery so I decided to stop here and consider the results of the 1st 3 Batteries and wait until the morning to see what the bank it at after a night of no charging/discharging.  I plan to shut the system down overnight (and put my food in cooler on the porch) and give it a full 6-hrs of non-use to get accurate readings.

Looking at those readings I have to say that it appears that after running on the charger for a few hours and then running on the equalize setting for 3 hours the batteries were still at only about 55% of cap.

The trojan doc I have says the following:  subtract 0.004 per 10 degrees below 80F to correct readings to 80F for voltage equivalents.

So looking at the above I'd have to subtract about 0.012 from each reading.

This: Batt 1:   1.230/1.230/1.232
Becomes this: Batt 1:   1.218/1.218/1.214 and roughly 55% of charge and a voltage of ~24.3v

Not really good for a system that had 5-6 hours of real charging and just finished 3 hours at 31v

I have to ask this now:  if I'm at only 55% of capacity now then should I calculate my capacity at 55% of 660AH (original AH rating of system)?  If so then I have only 363AH of cap and a usage of 35AH puts me close to 10% DOD.

However, in all honesty the bank appears to be all but dead at this point so perhaps I should bother looking at it that way and should just set the genset to come on at 23.5v under a load and hope that I can limp it along for a couple more months.

[MountainDon]

Equalizing before the majority of the cells have reached a fully charged state is not really EQ'ing. It's more like a bulk charge until the voltage reaches the EQ voltage set point.

If it was me, I would remove all the big loads and run the charger as long as it took to reach the absorb set point. Maybe I would even boost my absorb set point a little. Watch the temperature and the electrolyte level.

If the readings remain out of whack then, IMO, there is little point in trying to get the Trimetric calibrated. The batteries are simply working against you and the Trimetric.

I would do everything I could to ensure that one way or another the new batteries would not be run down to 50% or less and also ensure that they can reach full charge every normal day of use.

My all time favorite hydrometer has been the Hydrovolt for some time now. NAWS sells them at the best price I have seen. I have 2 of them now; one at the cabin and one at home. Automatic temperature compensation and very easy to read even with bifocals or whatever.

[OlJarhead]

OK after 8 hours of resting the battery bank went from 24.4v (I had not fully charged it since 3pm when I stopped the equalize) to 24.7v at 8am.  I then checked voltages and SG readings.

I realized afterwards that I should have charged the bank up but at least I had a clear picture of the condition of the batteries (I think) when I was done.  Here are the results:

As you can see I have some batteries with 'ok' readings and some that are clearly much lower.

My thought is that the equalize didn't do much for me but again, I failed to fully charge the bank before shutting down the power.  Perhaps that was a mistake but I think I can assume that the readings are accurate and the bank was at 75% of capacity by 8am.

It's clear the voltage under a light load drops about 0.3v from what it will rise to and I can assume that it's likely 1-2v under what it really is when under a load that I'd typically see (Fridge running, lights on etc).  This at least tells me that setting the generator auto start at 23.5v that it's probably closer to 24.5v without the load (the generator senses the drop to target voltage and waits to make sure it's not just a negative spike -- I believe it waits 2 minutes before starting at the target voltage.

My thinking is that if I take the best of the batteries and reduce the bank to 8 batteries then I'll have 440AH (if new and at 80F) and the 20Amps of charging I have should be able to handle that when it isn't in use.  When it is I have the generator to back up the solar.

My typical use of 80amps MAX in a day would put me close to 20% DOD and with the generator coming on around there it should hold up for a couple months I think.

So, what do you think?  My thought is to take the batteries with 1.22SG or better and use them and discard the ones below that.

I may also choose to get GCB's again as the money savings is tempting and with my new knowledge of batteries and my past 4+ years experience should give me more years of use though the trojans are still very tempting.

[OlJarhead]

 I noticed that the batteries which connected to the positive bus were the most abused.  I found this very interesting and wondered if I should reverse the strings?  I figured that if that battery took more abuse for some reason then by putting the best batteries there (reversing them) then perhaps I'd get more life out of the strings.  I also noticed that the two string with the lowest SG readings were the two original strings.

Either way it's clear the newer string may just be better and it's clear that the positive side of the bank draws down the battery nearest it the most for some reason.

Anyway, those are my thoughts.  I'm interested in any ideas and here's a question:

Should I conduct this experiment again by charging the batteries until the system floats and then kicking on the equalize setting to equalize, use the string until around 7PM then charge it again until 10PM so I know it's full and finally letting it rest overnight again?  I am not sure I can do this as I have to work tomorrow but I'm interested in your thoughts on that.

[nailit69]

Man... solar sounds complicated.  I'm gonna have to do some serious research/studying/reading on the subject.  I guess I thought it was a little more like the ronco cooker... "just set it and forget it" lol.

[OlJarhead]

It's not really complicated LOL just you can't abuse it and get away with it! LOL A well set up system with regular maintenance and no abuse should be pretty simple but I've made some mistakes.  For example, when I built my system I planned a system with 660AH at 12vdc and 615watts of charging.  The MPPT controller could take that and put 40amps into the battery bank (roughly) and the system should have managed that just fine.  But I abused that system and decided to add 2 more batteries which took me to 880AH @12v.  Now the system was a little under what I probably needed but again, had I been careful it would work.  I did however drain the bank a couple times when I was away by leaving something on I should not have.

Next I converted the bank to 24v and added 4 more batteries.  It was more than 3 years and you really should avoid adding batteries but I found 4 that were close in age and use (so I beleived anyway and I appear to have been mostly correct) and added them.  Now I had 660AH @24v and needed about 60Amps of charging but only had 20.  That's a bit of a problem but the backup generator can handle that.  However I made mistakes again and hereI am today looking at a rebuild.  My lessons learned:

1.  Size the bank for more than you think.
2.  Think in AMPS not WATTS.  It's Amps in vs Amps out pretty much. 
3.  Don't let the bank die EVER!  If you aren't there to protect it then don't use it or have the genset ready to back it up when you are gone (in my case I'm ok with that for the most part).

it's really not so bad and I enjoy having it.

[OlJarhead]

If I group the best 8 batteries by volts/SG to keep the relatively close to one another than the strings should look like this:

Of course that means one string will be better than the other but perhaps with some equalizing I can improve that a bit.

Thoughts?

I could take the last battery on the 2nd new string and swap it with the highest battery on the 1st string to keep them closer together but am not sure what affect that will have.  I'm also thinking of putting the better string first since the closest string to the bus (#3 in the first sheet) seems to take the most abuse.

[nailit69]

Thoughts???... not from me...

I'll have to pick your brain a little more when the time comes... I have no idea what any of this means.  Sounds reasonably simple in theory but my Dad is a battery freak (mostly cuz of RV'ing) he has all the hygrometers and digital analyzers and volt/amp meters and that yellow and green voltage table thingy and I still have never been able to grasp onto it.

[MountainDon]

Nailit69,  when you come to planning your system, you can do a couple things to make life easier and the batteries last longer.

Forget about a 12 volt system, unless you will absolutely for certain never want to use large draw items like microwave ovens,fridges, and so on. 12 volt battery systems may be good for a few lights and a TV. Go for a 24 volt system at least.

Do not put any batteries in parallel. Series only. That may mean you have to forgo using golf cart batteries and step up to something larger like L-16's. Or go to a 48 volt system with all batteries still in series.

Estimate power use as honestly as possible and do plan on use growing. It always does grow. Keep in mind that heating anything with electricity that comes from solar and batteries is quite expensive power.

Reduce power needs as much as possible. For  example, 120 VAC lighting LED's are now cheap enough to not bother with CFL's at all. Incandescent is, of course, out of the picture.


OJH,  I think dropping a string of the worst batteries is a good idea.  As for what batteries should go in each string, perhaps it would be best to try and make the voltage of one string match the other, at least as close as possible. That way the CC and inverter/charger might see voltage that is closer to the "truth". That's just theory.

And yes, I would try to charge the chosen batteries as much and as long as possible in order to try and get them as even as possible.

[OlJarhead]

Figured that might be best Don.  Thanks

On the topic of parallel batteries I did a bunch of reading articles by battery engineers and came to the conclusion that 2-3 strings are fine.  To be honest we run 2-3 strings in central office back up systems (HUGE systems that use 2v 3600AH batteries and 24 of them in series to get 48v -- incidentally we run -48v systems -- and 2-3 strings, possibly more I'd have to go look again to get the kind o reserve power a CO needs if the power goes out.).

I know many in the solar biz claim that you should only use a series setup but I'd argue that while that might be an excellent way to go it also has limitations.  For example, in a series parallel system you can shut down one back to work on it while the others continue to provide power.  This is something that happens all the time in the CO's.  You switch out one bank, get it back online and then switch out the next.  Means you're never out of power.

One thing I read though was that the wiring from each bank should be as close to exactly the same as possible.  Electricity takes the path of least resistance so having the inverter/charge controller/charger at one end of a bank and the strings laid out moving away from them isn't the best idea.  My idea for my new bank is to put all the main connections in the center of the bus and the battery strings the same distance from that point (one above, one below so to speak) so there isn't a difference.

So far I've left the bank alone since I have so much to get done but I think I'll leave it as is until my next trip out.  It seems to do fine when nothing is drawing off of it and the panels can keep it charged while I'm gone since the draw from the fans in the back is so small.

[OlJarhead]

Spoke with Bogart Engineering today on their meter and learned the following:

1.  The meter determines the system is at 100% when the P1 value (29.2v -- .4v under Absorption target voltage on MPPT controller) and the P2 Value (14amps -- 2% of battery capacity) -- this I knew but wanted to confirm.
2.  The capacity should be left at 660AH regardless of temperature. -- What he advised was that if you knew that it was 32F outside then 84AH of use would be about 20% of the battery bank capacity vs at 80F when it would only be 12%.  He suggested that if you only wanted to allow a maximum of 20% DOD then setting the system to kick the generator on at a shown 12% DOD in the winter would be better than leaving it until a full 132AH were drawn from the system (which would show 20%DOD on a 660Ah system).  I think this is key to understanding how to use the meter and is something I've struggled with.
3.  Capacity shown on the meter (in %) represents your overall AH cap minus all AH used on the since when not being charged.  So, as I suspected after I'd done a bunch of reading on the meter last weekend, a reading of 94% on a cap of 420AH (which I'd adjusted it to) meant I'd used 6% of 420AH

So in short, regardless of voltage on the system since you can't really trust it when under a load, it's best to watch the capacity percentage as a gauge of used AH's AND pay attention to voltage as a way of gauging battery health -- meaning that if the system uses only 25AH of capacity and is drawing down to 23.5v than something is clearly wrong since it really shouldn't drop much at all even on a 420AH bank.

[OlJarhead]

Came up with a cut list for my next stage of the deck.  I'll be adding another section that is 8 feet deep and 22 feet long then putting a roof over it.  My plan is to frame the roof with 3x6 rafters (due to the low headroom I'll have at the end in order to get some kind of decent slope to the roof.  I think I can get the slope better than 2:10, maybe 3:10 but it will be close and that will be with a 6'6" ceiling at the low end.  My dad is 6'4" so that will be a bit close but going higher means a much lower slope.  Perhaps I could consider deeper rafters and a lower slope though.  If I went with 2x8's rough sawn then I would have a stronger roof than the 3x6's I believe but I'd lose some height and go with less slope...this is my dilemma.

On a side note, the 2x8's would mean I'd get 4 out of each 8x8 cant which means they would mill faster whereas 3x6's would give me just 2 out of an 8x8 cant though with the 3x6's I'd also get one 2x8 out of the cant which I'd need for other things....hmmmm   

[OlJarhead]

Since deeper is better I'm now thinking that I'd be better off with 2x8's for rafters on the porch (vs 3x6's) and making the roof pitch shallower.  Looking at the span tables I should be fine with about a 2:12 pitch and spanning 10' on 24 centers.  These would be set on a ledger screwed to the cabin with ledger locks and then will rest on 6" D-Logs which will be set on 8+" log posts with log railing between them.  The log posts will be set at either 4' or 6' centers (haven't decided yet but 6' appears to be fine with the tables I've seen).  To do this I'd mill two 12' D logs 6" deep (width will be whatever the log width is but greater than 6" (most likely 8-10").  Then along the next section I'll run the same setup but the span will only be 8' since the deck is narrower on that side of the cabin (and around back will also be 8').

[OlJarhead]

My latest sketchup.  Started over as I wanted to clean up some lines but this is what I'm working towards now.

[OlJarhead]

This is what I'm aiming for

[OlJarhead]

Back to panelling in the loft but the going is SLOW

Seems putting up he panelling on the underside of a 12x12 pitched ceiling is tough work!  I've managed to get ONE row in so far but came down to check on the power.  I set the generator to come on at 24v so it will provide me with plenty of power but the battery bank is still at 24.6 despite running my miter saw and shop vac for every cut.  Takes a lot of power (I'm guessing about 2500 watts both combined) but the 4000 watt pure sine AIMS inverter is taking it just fine and the batteries are holding

[OlJarhead]

Flooding!  It's been in the low 40's during the day and the snow is melting off.  Rivers running down the driveways and roads out here.


Shot of the new fan (200mm) and hot water heater installed.


Blue LED's in the fans don't draw much power but add some light in the back room.


Some panelling I got done tonight (ceiling)

[OlJarhead]

Finished a couple more rows of panelling until I ran out of 4, 5 and 6" stock then got busy installing our 12v ceiling fan.  It's working nicely now and might also be ok on 24v since it has both written on the instructions (24v by hand, 12v by factory)...perhaps it's like the 24v Sureflo pumps that can run half speed on 12v.

Either way it's nice to have some air moving in the cabin now and it doesn't seem to use much power at all.

[nailit69]

I'm liking the porch.  We managed to make it up to our place this weekend... still 6-8" of snow on the ground but the higher sunny hills and south facing slopes are clear.  It was so much fun 

[OlJarhead]

I managed to get 4 trees felled and bucked into logs giving me (with one I'd previously dropped) about the 24 logs I needed for the new porch section and roof.  Most were dragged to the mill landing also though I still have a half dozen or so laying around the area and a lot of slash to stack and burn.

We'll be gone a while (longest in over a year) but should be back in early March to start milling.  My plan is to mill all the rafters, beams and floor joists and decking etc by mid april at the latest so I have have them dry for a while.  Then I'll start on the log posts and railings and with luck also get the D-Log beams set in place.  If I can get all that done by mid-may than I should be able to start framing the roof on the original deck and begin working on the foundation for the new section.

Needless to say I have A LOT of work to do this spring!  Need to have it done by July 10th if at all possible since I have a big family party that weekend at the cabin

[OlJarhead]

It's been forever since I've been to the cabin but I hope to be there in a week or so.  After our vacation in Italy we had to start packing up our home to sell as we'll be moving 80 miles closer to our cabin (moving to the town I have been working in and where my wife found work) so we're going to be stretched pretty thin between house hunting, selling our home and of course continuing to work on the cabin!

Life will be busy but I still hope to get my milling done and deck framed complete with roof -- I think I'm going to need some luck! lol