Temperature Coefficient Conversion %/C to %/K

sun_day
sun_day Registered Users Posts: 23 ✭✭
Hey all! I am using a software that only allow me to input %/K for the temperature coefficient. The solar module datasheet provided me a Voc temperature coefficient of -0.31%/C.

How can I convert %/C to %/K?

Thanks!

Comments

  • westbranch
    westbranch Solar Expert Posts: 5,183 ✭✭✭✭
    I think you are asking about *Kelvin, this might give you the answer...

    kel·vin

     (kĕl′vĭn)
    n. pl. kelvin Abbr. K
    1. A unit of absolute temperature equal to 1/273.16 of the absolute temperature of the triple point of water. One kelvin degree is equal to one Celsius degree. See Table at measurement.
    2. Kelvin A temperature scale in which zero occurs at absolute zero and each degree equals one kelvin. Water freezes at 273.15 K and boils at 373.15 K.


     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • sun_day
    sun_day Registered Users Posts: 23 ✭✭
    Hi, thanks for the response. I am asking the about %/C to %/K. Sorry if this is a simple conversion but I am not that good with it. Does this means it is just a direct conversion? As far as I know the -0.31 is in %
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    edited March 2016 #4
    sun_day said:
    Does this means it is just a direct conversion? As far as I know the -0.31 is in %
    A degree Kelvin is the same as a degree Celsius.  If the temp compensation is millivolts per degree K then it is also per degree C (no conversion necessary).  

    Unfortunately the temp coefficient is not linear with temperature... That's why the coefficient for your batteries is given as a percent per degree Kelvin.  In other words, the number of millivolts of compensation per degree depends on the temperature.   

    This is important because 30° C is 150% of 20° C, but 303° K (which is 30° C) is only 103% of 293° K (which is 20° C).  As a practical matter, the temperatures at which we operate our batteries is fairly narrow on the Kelvin scale and the temp compensation in millivolts per cell varies little over those temperatures. 

    Thus if the temp compensation were 4 millivolts per cell at 20 ° C, then it would be reduced by 3.1% (10 degrees X -.31% per degree) to 3.9 millivolts per cell at 30° C. 

    --vtMaps



    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • sun_day
    sun_day Registered Users Posts: 23 ✭✭
    Thanks for the reply. Does this means that -0.31%/C = -0.31%/K?
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    edited March 2016 #6
    I think so.  I am not too familiar with describing the temperature coefficient as percent per degree.  Consider that the temperature coefficient is also a function of the sulfuric acid concentration... that means that the coefficient changes with the SOC of the battery.  That means that as the battery charges (even if it stays at the same temperature), the temp coefficient is changing.

    I would prefer to set a temperature compensation of millivolts per degree.  In the temperature range we run our batteries, that approach is good enough.  Ultimately, what matters is the SG of your batteries.... if the batteries are cold and the SG is not up to spec, increase your absorb time and/or absorb voltage and/or your temp compensation.   If the batteries are hot and the SG is up to spec, but you are using a lot of distilled water, then reduce your absorb time and/or absorb voltage and/or increase your temp compensation

    --vtMaps

    EDIT: I just reread this thread and realized you are asking about the temp coefficient of Voc of your modules.  I have been writing about temp coefficient of battery charging.   What sort of software are you using that needs the temp coefficient of Voc?
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • sun_day
    sun_day Registered Users Posts: 23 ✭✭
    edited March 2016 #7
    Hello,

    Thanks for the reply. I am using Tritec TRI-KA software (it is an IV curve equipment and software). I need to input the temperature coefficient from the datasheet but the units they have in the software is only in %/K (or mV/K). However, the solar module datasheet provided the temperature coefficient only in %/C.

    Therefore I don't think I can directly input -0.31%/C into the software as the unit in the software is %/K.

    Really at a lost here... T_T
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    sun_day said:
     the units they have in the software is only in %/K (or mV/K). However, the solar module datasheet provided the temperature coefficient only in %/C.
    I can't help you with the percents... not familiar with that convention. 

    However a K degree and a C degree are the same size, therefore mV/K is the same as mV/C.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    Percent is simply multiplied x Voc of panel. If you have Vmp=21 volts, use 21 volts. If panel is 40 volts, use 40 volts. If testing a cell, then ~0.6 volts.

    If mVolts/C -- The value will change based on Voc of cell/panel. (mV/C will be different when Voc is different between test subjects).
    • Voc * %/C = mV/C
    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • sun_day
    sun_day Registered Users Posts: 23 ✭✭
    Thanks guys for the reply.

    @Bill - I don't think the formula finds the %/C to %/K as I think you are trying to show how to convert %/C to mV/C.

    Does anyone else still have any idea on the unit change from %/C to %/K? @_@
  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    edited March 2016 #11
    Delta (K) = Delta (C)

    If you are doing differences in temperatures.

    Absolute temperature conversion:

    °C = K - 273

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • vtmaps
    vtmaps Solar Expert Posts: 3,741 ✭✭✭✭
    sun_day said:
    Does anyone else still have any idea on the unit change from %/C to %/K? @_@
    Bill explained that the percent (per degree) multiplied by the Voc gives millivolts (per degree).    Since both degrees (K and C) are the same size, it doesn't matter whether you use Kelvin or Celsius.... the coefficient is the same for both.

    --vtMaps
    4 X 235watt Samsung, Midnite ePanel, Outback VFX3524 FM60 & mate, 4 Interstate L16, trimetric, Honda eu2000i
  • sun_day
    sun_day Registered Users Posts: 23 ✭✭
    Thanks! I get it now. I will convert it to mV to find. Does this applies to Isc too?
  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin
    What exactly are you asking about Isc? Yes, there is a temperature correction for Isc (and Imp). As far as I know, Isc and Imp will rise a little bit with increase in temperature (less than the correction for Vxx vs temperature).

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • sun_day
    sun_day Registered Users Posts: 23 ✭✭
    Hi guys,

    Thanks lot for the help and teaching. I have understood it now.
  • AGTM_Doc
    AGTM_Doc Registered Users Posts: 2
    edited March 2019 #16

    Hello,

    A module I am working with shows the Temp Coeff for Voc as -(80±10)mV/°C How do I convert this to %/C?

    Thank you,

    Ed

  • BB.
    BB. Super Moderators, Administrators Posts: 33,433 admin

    Here is an example for a Voc~21 volts ("12 volt" solar panel)

    • 80 mV / °C = 0.080 volts / °C = (I think) 0.080 volts per panel per °C = 0.080 volts / (Voc panel * "delta" °C )
    • (0.08 volts per °C) / 21 Volts Voc = 0.00381 per °C = 0.381% per °C

    Note we are missing one variable, the per "what" device ... I am guessing this is per solar panel (i.e., per 21 Voc at standard temperature).

    -Bill

    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • essie94
    essie94 Registered Users Posts: 1
    edited August 2022 #18
    A typical value for crystalline modules is βOC= 0.34 %/°K (the coefficient is usually given per °Kelvin, not per °Celsius).

    However, as there is a linear relationship between the Kelvin and Celsius scales, the use of either will lead to the same result.

    If the temperature goes up to 35°C, the difference to STC is 10°K.

    With the given coefficient the open-circuit voltage will go down by 3.4%.

    For a roof-top installation, a module temperature of more than 50°C can be expected.

    As a rule of thumb, the open-circuit voltage of an irradiated module can be calculated as Voltage at STC -10%.


    -Es
  • JonBruce
    JonBruce Registered Users Posts: 1
    edited February 2023 #19
    I'm aware the initial post is quite old, but I have potentially valuable information to add. 

    As for the OP, an increase of 1°C = an increase of 1°K, therefore the temperature coefficient should be the same as it is per degree. However, for F->C or C->F conversion, this is not at all a simple or linear thing, as the two do not scale linearly with each other. 

    In my case, the module I am using has their temperature coefficient listed in F, at -0.39%/°F. To convert to C, I ran a calculation at an assumed 40V @ 70°F and calculated that I would have 38.44V @ 80°F. I then calculated 70°F = 21.11°C, and that 31.11°C = 88°F. Note the common factor being an increase of 10° in either F or C. Also, be aware that decimals are being rounded to no more than 2 digits after the decimal. I then calculated the voltage at 88°F(31.11°C) = 37.192V. I then found the voltage change in 10°C = 2.808V, which when divided by the initial 40V shows -7.02% change, or -0.702%/°C. Upon testing this across a range of numbers and calculating back to F and re-running the calculation of voltage change, I have found it to be consistently accurate. This is how I successfully converted %V change/°F to %V change/°C. 

    That's the proof, the example showing it works. The formula is much, much simpler as it turns out. 

    V%/°C = (V%/°F) x (9/5)
    V%/°F = (V%/°C) x (5/9)

    Examples:
    V%/°C = -0.390x9/5 = -0.702
    V%/°F = -0.702x5/9 = -0.390