Cooling the batteries

2

Comments

  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    edited September 5 #32
    Softdown,     keep in mind that the 1st two days were using low DODs (about 11% so not a hard charge (controllers only harvested 12 kwhs today as the house AC mini-split burned power).     Tomorrow will be a much tougher test with me pulling the batteries down at least 23% DOD so a lot more bulk heat added charging tomorrow.       We ran 3 loads of laundry tonight and the AC until 11 so the batteries will be a tad lower than average.

    Pulling fresh well water the water temp is 72-76 degrees during the summer.     Well is 390' deep but the pipe runs 18" underground 70 yards from the well picking up ground heat ( summer temps here in my part of GA).

    But draining and refilling the box daily (20 minutes) with cooler well water also uses a lot of pump electricity but may be a good choice (see test results in post #27).     Instead I'm trying to leave the water in the battery box with weekly replacement to avoid pumping power costs while advoiding scum and algae problems.     I've started using a $600 1/4hp chiller to cool the water using excess power from the panels during absorb and float.         

    It is now 11pm.      Chiller cut off at 6.      Current ambient temp in garage (just closed the big doors) is 81 and batteries in water are 74 degrees.

    Only day 2 using this water bath and chiller system but so far it's doing better than I expected.       The chiller is only consuming electricity I would have lost when the batteries were in absorb and float since I can't sell to the grid.

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • mcgivormcgivor Posts: 2,203Solar Expert ✭✭✭✭✭
    What I discovered is it takes a significant amount of time to get the heat out of the mass of lead, once stable the swings are reduced, cladding  the vessel with styrofoam insulation helps losses and condensation, I used a plastic container so insulation value is less than plywood. The spaces I had between the batteries and walls were larger than yours, I used styrofoam packing pellets to float on the surface, my biggest issue was as the hottest months are in what it's technically winter, it's also extremely dry, rodents, geckos and ants discovered a new source of water, sometimes dying to get a drink.
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery bank 

  • littleharbor2littleharbor2 Posts: 1,027Solar Expert ✭✭✭✭
    Do you have a temp sensor on the batteries? If so does it get submerged?

    2.1 Kw Suntech 175 mono, Classic 200, Trace SW 4024 ( 15 years old  but brand new out of sealed factory box Jan. 2015), Bogart Tri-metric, 700 ah @24 volt AGM battery bank. Plenty of Baja Sea of Cortez sunshine.

  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    Ouch, providing the local water source for pests was something I hadn't considered.

    While not shown in the picture the box has a lid that is usually left open to facilitate heat escaping.       There is also an exhaust fan.      You can see the 2 air intake holes in the box which are covered with heavy screen.       Those intake holes double as an emergency overflow release.      I never wanted to risk the water getting overly full and into the outer battery cases or under the rubber liner so those holes are a bit lower than the top of the pond liner and battery tops.

    Since the box needed to be raised off the floor to allow for a drain the box has 2x4s insulating it from the concrete.       The boards are spaced so if I ever get water under the box it can get some airflow and dry out. The outer wood band surrounding the box like a belt is to reinforce against the water pressing out on the sides.

    I wanted to make the box more square and stay out from under the electronics but putting the batteries in 2 rows would have made maintaining the batteries a monthly torture having to reach over the 1st row of batteries to reach the 2nd row. 

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    Do you have a temp sensor on the batteries? If so does it get submerged?

    Rolls 5000 series batteries have a double case.      The three 2v cells are in an outer case which has 1/2" deep ridges inside so there are large areas where the inner battery is air gapped from the outer shell.       The sensor is directly attached to the inner 2v cell and protected from the water by the outer shell.      Space between the inner and outer shells means the sensor does not touch the outer shell (3/8" air gap between the sensor and outer shell) and is less affected thermally by the cool water outside the outer shell.

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • mcgivormcgivor Posts: 2,203Solar Expert ✭✭✭✭✭
    Stainless steel mosquito mesh is what I use to prevent ingress of unwanted guests, mine was purely an experimental quest, moving to LFP largely eliminates many of the issues of cooling, but I'm sure when temperatures rise into the 40°C ambient   range,  I'll  revisit  some sort of cooling. 
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery bank 

  • softdownsoftdown Posts: 1,889Solar Expert ✭✭✭✭
    That space between the inner and outer shells lowers the effectiveness of the cooler water by a significant amount as you know. What would happen if you drilled a lot of circulation holes in the outer shells?

    Seems like you are now leaving the chiller on for much longer than you were initially. I keep wondering about the net effect if the chiller started working before the batteries were sent into charge mode. A greater DOD but does a very short period with a higher DOD effect the battery longevity as much as an overnight period of higher DOD? 

    A 1/4 hp chiller is working well enough for you. I would be interested in seeing how a smaller chiller working on a set temperature and working 24/7 would do. Say a 1/6 hp chiller set to keep the water below 70F. If the water was always below 70F the battery temperatures should follow. Also a long term study on employing the 1/4 hp chiller on a set temperature and working 24/7. Or maybe closer to 16/7. Or variable with the seasonal sun output. So many variables to study. 

    Amazing that such variables do not seem to have been studied. Or the results are virtually unavailable. A company like Outback or Schneider could fund such a test and make chillers and cooling boxes available. The resulting "complication" would likely scare away many of the less scientifically minded. 

    Too bad your well is moderately deep. A holding tank would soon reach ambient air temperatures and be almost useless. 

    Hope this all makes sense. It is mostly speculating on the most efficient methodology to employ a chiller in a hot environment. Give us an inch, we want a foot!
    First Bank:16 180 watt Grape Solar with  FM80 controller and 3648 Inverter....Fullriver 8D AGM solar batteries. Second Bank/MacGyver Special: 10 165(?) watt BP Solar with Renogy MPPT 40A controller/ and Xantrex C-35 PWM controller/ and Morningstar PWM controller...Cotek 24V PSW inverter....forklift and diesel locomotive batteries
  • jonrjonr Posts: 1,074Solar Expert ✭✭✭✭
    edited September 5 #39
    IMO, the optimal design is a large chiller and a large water tank such that all the water cooling can be done when you have excess power and there is no charge effect on the batteries. 

    Waterwheel's numbers indicate that there would be some benefit to more water. 
  • Dave AngeliniDave Angelini Posts: 4,010Solar Expert ✭✭✭✭
    Too bad you can't cool the electronics also with some evap cooling. Too humid for that....Still a good project!
    "we go where power lines don't" Sierra Mountains near Mariposa/Yosemite CA
     http://members.sti.net/offgridsolar/
    E-mail [email protected]

  • softdownsoftdown Posts: 1,889Solar Expert ✭✭✭✭
    Imagine a huge "refrigerator" with 3" - 4" foam walls. Housing batteries and electronics. Would be much like a walk in freezer. 

    One might consider directing some button fans over the cooled water and towards the electronics. Too bad about the humidity in Georgia. 
    First Bank:16 180 watt Grape Solar with  FM80 controller and 3648 Inverter....Fullriver 8D AGM solar batteries. Second Bank/MacGyver Special: 10 165(?) watt BP Solar with Renogy MPPT 40A controller/ and Xantrex C-35 PWM controller/ and Morningstar PWM controller...Cotek 24V PSW inverter....forklift and diesel locomotive batteries
  • softdownsoftdown Posts: 1,889Solar Expert ✭✭✭✭
    jonr said:
    IMO, the optimal design is a large chiller and a large water tank such that all the water cooling can be done when you have excess power and there is no charge effect on the batteries. 

    Waterwheel's numbers indicate that there would be some benefit to more water. 
    That seems like the intuitive solution. On the other hand, refrigerators with smaller compressors use less net energy. 

    I am amazed that it is 2018 and would seem to be "breaking ground" on cooling solar batteries with a chiller. 

    I also remember that water chillers tend to use a lot of energy in a reef tank, thus releasing a lot of heat energy to the room. That heat energy somehow needs to be directed outside if possible.

    Imagine the consternation of WaterWheel at this point. We should probably all be congratulating him on progress to date.

    I am hardly ever satisfied for more than a couple seconds...
    First Bank:16 180 watt Grape Solar with  FM80 controller and 3648 Inverter....Fullriver 8D AGM solar batteries. Second Bank/MacGyver Special: 10 165(?) watt BP Solar with Renogy MPPT 40A controller/ and Xantrex C-35 PWM controller/ and Morningstar PWM controller...Cotek 24V PSW inverter....forklift and diesel locomotive batteries
  • jonrjonr Posts: 1,074Solar Expert ✭✭✭✭
    True, congrats to Waterwheel for having an effective, affordable system for increasing LA battery life.  

    >refrigerators with smaller compressors use less net energy

    Given any small compressor, one can design a larger compressor that is equally efficient.  But it will be more expensive.

    Would be interesting to discuss Winter cooling.  If the batteries are normally 75F, is it worth it to use a little outside air to bring them to 60F?

  • BB.BB. Posts: 27,883Super Moderators, Administrators admin
    The engineering formula is 2x increase in life for every 10C/18F reduction in temperature (or 1/2 the life for every 10C increase).

    Note that if your batteries are only "hot" for 12 hours per day for 3 months of summer--That means instead of aging at a rate of 1.5 months during elevated temperature, they are aging 3 months (or 3 months of summer is 1.5 normal + 3 elevated temp aging = 4.5 months of "aging" at 10C increase in temperature for 1/2 day) over those 3 months of summer or ~13.5 months aging in 12 months of year.

    If I got the math right (no guarantees).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    jonr said:
    IMO, the optimal design is a large chiller and a large water tank such that all the water cooling can be done when you have excess power and there is no charge effect on the batteries. 

    Waterwheel's numbers indicate that there would be some benefit to more water. 

    Actually my current numbers indicate that there wouldn't be much advantage to using a large tank of pre-chilled water and I may have been able to get away with a 1/6 hp chiller.        The chilled water holds well through the night.
    The chiller has been off since 6pm yesterday.      At 7:30 am today the batteries were 73 and the air in the garage was 79, I didn't check water temp but I would assume about 73 degrees, same as the batteries.     The batteries went down to 76% SOC so we're getting a good hard charge today.     Cloudy earlier but 84 amps going in at noon when the controller went into float yet the batteries only went up to 77 degrees.      The chiller timer runs the chiller from noon to 6 which is when the heating is the worst but as the days get shorter I'll shut the chiller down earlier.

    I'll post the days numbers at 6 after the sun is mostly off the panels.

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    Too bad you can't cool the electronics also with some evap cooling. Too humid for that....Still a good project!

    While I'd prefer not to have the batteries under the inverter I figures having a 65-75 degree tank 18" below the inverter and controllers fan's intake may help cool the electronics a tiny bit.

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • Dave AngeliniDave Angelini Posts: 4,010Solar Expert ✭✭✭✭
    BB. said:
    The engineering formula is 2x increase in life for every 10C/18F reduction in temperature (or 1/2 the life for every 10C increase).

    Note that if your batteries are only "hot" for 12 hours per day for 3 months of summer--That means instead of aging at a rate of 1.5 months during elevated temperature, they are aging 3 months (or 3 months of summer is 1.5 normal + 3 elevated temp aging = 4.5 months of "aging" at 10C increase in temperature for 1/2 day) over those 3 months of summer or ~13.5 months aging in 12 months of year.

    If I got the math right (no guarantees).

    -Bill
    I think you do have good numbers!  I also think waterwheel has some of the best electronics in the world for heat. I still have one in an ambient shed south of Tucson that has been flawless since 2007. Lots of different batteries over that period....
    "we go where power lines don't" Sierra Mountains near Mariposa/Yosemite CA
     http://members.sti.net/offgridsolar/
    E-mail [email protected]

  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    edited September 5 #48
    BB. said:
    The engineering formula is 2x increase in life for every 10C/18F reduction in temperature (or 1/2 the life for every 10C increase).

    Note that if your batteries are only "hot" for 12 hours per day for 3 months of summer--That means instead of aging at a rate of 1.5 months during elevated temperature, they are aging 3 months (or 3 months of summer is 1.5 normal + 3 elevated temp aging = 4.5 months of "aging" at 10C increase in temperature for 1/2 day) over those 3 months of summer or ~13.5 months aging in 12 months of year.

    If I got the math right (no guarantees).

    -Bill

    Bill,      More like 6- 7 months out of the year that the batteries spend a lot of time at over 80 degrees where I am.
    Three months of the year I rarely see the batteries below 88 degrees in the morning and bumping over 100 most afternoons.

    I suspect getting the batteries a few inches off the concrete floor is also helping them cool overnight.       The box had to be raised so the water drain could be installed.

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • softdownsoftdown Posts: 1,889Solar Expert ✭✭✭✭
    BB. said:
    The engineering formula is 2x increase in life for every 10C/18F reduction in temperature (or 1/2 the life for every 10C increase).

    Note that if your batteries are only "hot" for 12 hours per day for 3 months of summer--That means instead of aging at a rate of 1.5 months during elevated temperature, they are aging 3 months (or 3 months of summer is 1.5 normal + 3 elevated temp aging = 4.5 months of "aging" at 10C increase in temperature for 1/2 day) over those 3 months of summer or ~13.5 months aging in 12 months of year.

    If I got the math right (no guarantees).

    -Bill
    No problem with the figures. May have a problem with the formula. Georgia is a pretty large state and does have some part of Appalachia. Going from an "average location in Georgia" however, I would expect closer to 4-5 months of troublesome heat. 

    For example, I literally rained sweat while doing roofing work in S Florida in very early April. It was hot in April and will likely continue to be "hot" through October. So...seven months of problematic heat in S. Florida. May equal 4-5 months of similar heat problems in Georgia. Or ~ 14.5 months aging in 12 months. 

    Above just stated that lots of batteries have been exhausted in an ambient Tucson location since 2007. I will assume that means at least three sets of batteries, or more, in 11 years. Tucson is presumably hotter than WW's locale of course. 
    First Bank:16 180 watt Grape Solar with  FM80 controller and 3648 Inverter....Fullriver 8D AGM solar batteries. Second Bank/MacGyver Special: 10 165(?) watt BP Solar with Renogy MPPT 40A controller/ and Xantrex C-35 PWM controller/ and Morningstar PWM controller...Cotek 24V PSW inverter....forklift and diesel locomotive batteries
  • BB.BB. Posts: 27,883Super Moderators, Administrators admin
    I was just trying to make an example of how to use math to do a cost benefit analysis... Not attempting to quantify any specific installation.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    edited September 11 #51
    Using your formula and considering the 77 degree temp most battery manufacturers use  I see my batteries aging an extra 4-5 months per year.       Of course during the winter that aging process probably slows below the average expected aging rate at 77 degrees but let's say that because of heat my batteries age 16.5 months for every year (4.5 months well over 88 degrees. and often over 100 degrees).      That means every 5 years the batteries go through an additional 22.5 months of aging (4 years worth of excess heat related aging if they last 11 years).      

    After freight charges I'm using a $10,900 set of batteries (hurts to even type that big number).     That's $1090 per year for batteries if they last the 10 year warranty period or $5450 for 5 years.       Assuming the $600 water chiller and pump lasts 5 years but saves about 2 years worth of heat aging that means the $600ish investment is saving me roughly $2200 over a 5 year period in addition to delaying the wonderful job of replacing those 340 lb batteries for an additional 4 years (theoretically).

    I understand that these are all theoretical numbers but we've all seen how batteries in Canada last a lot longer than batteries in Tucson AZ.       
    Using your much lower 1.5 months additional heat related damage means I'd save about $75 over 5 years when you conceder chiller cost vs battery wear cost so I'd say that your 1.5 months lost life per year of service is about the break even point as to whether or not using a chiller is financially worthwhile.

    Fun fact,       In an effort to make putting the batteries in the water box easier I attached the pond liner and front wall of the box in such a way that the front wall of the box can be folded down flat on the floor in front of the box.        This allowed us to not have to heave those batteries up and over to get them into the water box.

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    edited September 5 #52
    Day 3 results

    76% SOC this morning, light and patchy cloud much of the day so I cut off the chiller at 5:45 since the clouds were making it hard to hold float with the 7 amp draw from the chiller but I did manage to get a full charge today and sit in absorb the last 90 minutes.

    8am   battery 73 degrees, air 79 degrees
    6pm   battery 75 degrees, air 91 degrees, water 64 degrees

    This is the last daily result post although I'll check in occasionally to update.

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • jonrjonr Posts: 1,074Solar Expert ✭✭✭✭
    edited September 5 #53
    Sandia reports "every 8°C (15°F) rise in temperature will cut the battery life in half."

    https://prod.sandia.gov/techlib-noauth/access-control.cgi/2004/043149.pdf

    77F is used as a reference temp, but AFAIK chemically there is nothing special about it and even cooler is better for life (ie, there is no "above this is a problem" point).  So many may be able to cool 15F below normal the entire year, making the math simple.
  • WaterWheelWaterWheel Posts: 282Registered Users ✭✭✭
    I'm also seeing some articles about temps over 88 degrees affecting lithium battery life.        I'm sure it's not as sever as with LAs but something to conceder if in a hot area. 

    Conext XW6848 with PDP, SCP, 80/600 controller, 60/150 controller and Conext battery monitor

    21 SW280 panels on Schletter ground mount

    48v Rolls 6CS 27P

  • VicVic Posts: 2,906Solar Expert ✭✭✭✭
    edited September 5 #55

    IMO,  for LA batteries,   the battery manufacturer chooses the Reference Temperature  --  often it is 77F,   sometimes 80F.

    It is the Reference Temperature that is used for Capacity data,   etc.

    Many battery manufacturers do seem to use the "for every 10 degree C increase in battery temperature,  above reference temperature,  battery life halved",   and the converse.

    Had not seen the Sandia 8 C increase,   battery life is halved.

    It has seemed that the 10C number was based on chemical activity,   in general.

    FWIW,   IMO,   Vic


    Off Grid - Two systems -- 4 SW+ 5548 Inverters, Surrette 4KS25 1280 AH [email protected], 11.1 KW STC PV, 4X MidNite Classic 150 w/ WBjrs, Beta KID on S-530s, MX-60s, MN Bkrs/Boxes.  25 KVA Polyphase Kubota diesel,  Honda Eu6500isa,  Eu3000is-es, Eu2000,  Eu1000 gensets.  Thanks Wind-Sun for this great Forum.
  • BB.BB. Posts: 27,883Super Moderators, Administrators admin
    The actual number can be calculated using calculus (yea--I had to clunk through the calculations in Calc class many decades ago)--There is a common activation value used for most materials--And that works out very closely 10C and 1/2 (or 2x) change in "life/aging".

    https://www.electronics-cooling.com/2017/08/10c-increase-temperature-really-reduce-life-electronics-half/

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • softdownsoftdown Posts: 1,889Solar Expert ✭✭✭✭
    Might be something to consider though probably not. Costco sells 20 pounds of ice for $2 - in Colorado. They sell no ice at all in Florida where people just might need some. 

    Seeing what 20 pounds of ice does would be worth at least $2.
    First Bank:16 180 watt Grape Solar with  FM80 controller and 3648 Inverter....Fullriver 8D AGM solar batteries. Second Bank/MacGyver Special: 10 165(?) watt BP Solar with Renogy MPPT 40A controller/ and Xantrex C-35 PWM controller/ and Morningstar PWM controller...Cotek 24V PSW inverter....forklift and diesel locomotive batteries
  • mike95490mike95490 Posts: 7,901Solar Expert ✭✭✭✭
    softdown said:
    Might be something to consider though probably not. Costco sells 20 pounds of ice for $2 - in Colorado.
    Wow, I just bought 10lbs for $22 at grocery store.  I'll have to see if my local costco carries dry ice
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • BB.BB. Posts: 27,883Super Moderators, Administrators admin
    $2 for 20 lbs is (as far as I know) for "water ice".... Not "dry ice".

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mike95490mike95490 Posts: 7,901Solar Expert ✭✭✭✭
    BB. said:
    $2 for 20 lbs is (as far as I know) for "water ice".... Not "dry ice".  -Bill
    ahh  still in burning man mode
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • softdownsoftdown Posts: 1,889Solar Expert ✭✭✭✭
    Think I'm in burned mode. 2018 has been thoroughly unbelievable. Been nine months of a "Rocky" fight.
    First Bank:16 180 watt Grape Solar with  FM80 controller and 3648 Inverter....Fullriver 8D AGM solar batteries. Second Bank/MacGyver Special: 10 165(?) watt BP Solar with Renogy MPPT 40A controller/ and Xantrex C-35 PWM controller/ and Morningstar PWM controller...Cotek 24V PSW inverter....forklift and diesel locomotive batteries
2
Sign In or Register to comment.