[RE-wrenches] Cable Sizing - revisited, Ambient Temp

[Dave Click]

John,
240.4(B) allows the ampacity of a wire to drop below the OCPD rating so 
you could argue that wouldn't be a safety issue. It's always irritated 
me that these NEC derating factors have two decimal places so that you 
feel your calculations are extremely exact... and then your 
carefully-derated-to-126.39A wire has no problem at all on a 150A 
breaker. But that's the way it is.

It takes a lot of heating to exceed the 90C rating of a wire, but yes 
theoretically that would damage the insulation and cause a fault. As for 
voltage drop of your wires, I don't think that's as big an issue as you 
think. Heating from 40C to 80C, a wire is going to see its resistance 
increase by about 15%, so your voltage drop may go from 1.5% to 1.73%. 
Your module voltage drop is going to be the more pressing concern.

If on your drawings you state something like "conduits to be minimum 3.5 
inches above the roof" (or for residential, put the run in the attic) 
then that puts you at a 17C adder which is pretty manageable for most 
areas of the US (worst case, you're in the 61-70C temp derate range). 
And again, in some cases where you have a very hot section of conduit, 
the 10'/10% rule may let you ignore that localized heating. Since you're 
the one stamping these drawings, you have to stay within your comfort 
zone- add another 10C if you're worried about it and add some extra 
expansion joints. Or you can specify that your contractors shade all 
rooftop conduit, but I imagine that would limit your repeat business.

Dave

John Wadley wrote:
> Dave,
> Thanks for responding in Mr. Brooks place. Since ASHREA 2% is not the
> very worst case, it seems like it might be possible for the ampacity of
> the wire chosen to dip below the rating of the OCPD protecting it, if
> there is not much margin. I've been trying to rationalize whether this
> would become a safety issue. I don't think it would since the OCPD
> protects the wire from a current source increasing beyond the expected
> design output. I don't think there is much chance of that for a PV
> module (unless there was a short between two strings). I think the
> increased heating would more likely increase wire resistance/voltage
> drop and lower production. With enough voltage drop, the inverter might
> shut off.
> I guess my new concern is in the most severe case where there is solar
> concentration on a short section of conduit. Here, the heating effect of
> both the elevated ambient temp and reduced wire resistance might lead to
> premature failure of the wire insulation. If the combined heating
> effects exceed the 90C rating of the wire, does the insulation embrittle
> or melt? In either case, I foresee a grounding fault, and if the GFCI
> failed, it could spark a fire.
> I know the best solution is to keep conduit shaded and avoid these worst
> case solar concentrating conditions. Sometimes. when I design a system
> for a new contractor, I don't always know exactly where they plan to run
> conduit on a roof (nor can I control it) and I start "what-if'ing"
> whether my design numbers will be conservative enough to prevent a
> system failure or a fire.
>
> Thanks and regards,
> John Wadley, PE
> NABCEP Certified Solar PV Installer (TM)
> Wadley Engineering
>
>  > Date: Tue, 25 Jan 2011 15:37:20 -0500
>  > From: Dave Click <daveclick at fsec.ucf.edu>
>  > To: re-wrenches at lists.re-wrenches.org
>  > Subject: Re: [RE-wrenches] Cable Sizing - revisited, Ambient Temp
>  > Message-ID: <4D3F3480.9030606 at fsec.ucf.edu>
>  > Content-Type: text/plain; charset=UTF-8; format=flowed
>  >
>  > John,
>  >
>  > The 2% "ambient temperature" from ASHRAE is the appropriate starting
>  > point to use for these calculations. For some additional background I'll
>  > quote Bill:
>  >
>  > **
>  > ASHRAE bases its ?warm?season temperature conditions? for each city on
>  > annual percentiles of 0.4%, 1.0% and 2.0%. As an example, the June 2.0%
>  > dry?bulb design temperature for Atlanta is 91.7?F. Therefore, based on a
>  > 30?day month (i.e. 720 hours), the actual temperatures can be expected
>  > to exceed 91.7?F a total of 14 hours a month. The corresponding 1.0%
>  > design temperature (93.1?F) can be expected to be exceeded for 7 hours a
>  > month; while the 0.4% design temperature (94.6?F) can be expected to be
>  > exceeded for 3 hours a month (column 2).
>  > **
>  >
>  > In Jim Dunlop's example it sounds like he's starting with the summer
>  > ambient high (likely around 90F / 32C) and adding the 310.15(B)(2)(c)
>  > 33C figure to reach the 61-70C range.
>  >
>  > IMHO, ASHRAE 2% high temperature should be the "standard practice" for
>  > these conditions when calculating your base ambient temperature before
>  > additional adders. There are going to be site-specific conditions like
>  > your example where the conduits may heat up more than 310.15(B)(2)(c)
>  > requires; in that case I think you'd be on the right track to make your
>  > own field measurements to determine an appropriate temperature. In some
>  > cases the 10%/10ft rule may mean you can ignore short hot spots in the
>  > wire. If you had a situation where the conduit was in direct sunlight,
>  > plus light was being reflected off the roof and a light-colored wall
>  > behind the conduit, I suppose that would yield more heating than what
>  > the CDA study had found (http://www.iaei.org/magazine/?p=1743). I'm not
>  > sure that shooting an IR thermometer is the best option here; if you
>  > want to best replicate the study conditions, you may put a conduit
>  > section up on the roof in the desired location, put a temperature sensor
>  > in the conduit, and let it soak. Then compare that number with [ASHRAE
>  > 2% + 310.15(B)(2)(c)] and pick the higher number. Or just add an
>  > additional 10C on top of ASHRAE+B2c and be done with it...
>  >
>  > I think your plan of trying to measure the temperature 4" off the roof
>  > where the conduit sits, and then adding the additional 17C (or whatever)
>  > from the Table, will be too conservative; your initial measurement will
>  > be affected by some of the heating that's wrapped into the
>  > 310.15(B)(2)(c) factor and you'd be double-counting that effect.
>  >
>  > Hope that helps.
>  > Dave
>  >
>  > -------- Original Message --------
>  > Subject: [RE-wrenches] Cable Sizing - revisited, Ambient Temp
>  > From: John Wadley <wadleyjk at hotmail.com>
>  > To: RE-wrenches <re-wrenches at lists.re-wrenches.org>
>  > Date: 2011/1/22 02:40
>  >
>  > > Mr. Brooks,
>  > > You replied to Mr. Parrish back in 2009 with this example (below) on
>  > > properly applying all the deratings to ampacity for wire sizing. I have
>  > > a bit of confusion and a question about the definition of "ambient
>  > > temperature". You define it below as the ASHREA 2% high temp. My NEC
>  > > 2008 (310.15 (2) Except No. 5 (3) (b) FPN) mentions it being an
> "average
>  > > ASHREA" number. The only definition for "ambient temp" in NEC I could
>  > > find was 310.10 FPN (1) which says it "varies along the length of the
>  > > conductor by time and place". In the Photovoltaic Systems by Dunlop, p.
>  > > 288, he cites a sizing example without saying where ambient comes from
>  > > but uses 61C-70C (142F-158F) derate factor (0.58) for a sunlit roof top
>  > > conduit. He does not say how he arrives at that tempature range, but I
>  > > suspect he started with the 90F rating of the USE-2 conductor in the
>  > > example and added a Table 310.15 (B) (2) (c) adder of 33C. Other
>  > > articles I've read talk about conditions like an unventilated attic
> or a
>  > > sunlit jbox on a roof where ambient temps could reach 150F. I can also
>  > > think of a situation where on a flat roof with a surrounding parapet
>  > > wall, the sunlight shining into a corner would act like a solar oven on
>  > > any conduit running close to the corner. So, given all these
> definitions
>  > > and possible exceptions to the definition of "ambient temperature",
> does
>  > > your original definition (ASHREA 2% high temp) still stand as standard
>  > > practice for most conditions and are there situations where one should
>  > > use something other than that defined value? If one is unsure of an
>  > > exceptional situation, would it make sense to use an IR thermometer to
>  > > measure free air temp on a sunny, calm day and then the air temp
> exactly
>  > > where conduit might run and use the temp delta as an adder (like Table
>  > > 310.15 (B) 2 (c)) to the ASHREA 2% high temp to arrive at a new,
>  > > situational ambient temp before applying the other factors cited?
>  > >
>  > > Thanks in advance,
>  > > John Wadley, PE
>  > > Wadley Engineering
>  > > NABCEP Certified Solar PV Installer (TM)
>  > > Dallas, TX
>  > >
>  > > Peter,
>  > >
>  > > We cannot use load diversity to increase the number of conductors
> in a PV
>  > > conduit since there generally is little diversity among the conductors,
>  > > particularly on large arrays.
>  > >
>  > > The more traditional conduit adjustment table to use is Table
>  > > 310.15(B)(2)(a). The value from this table is multiplied by the
> temperature
>  > > adjustment factor in Table 310.16. The key is what to use as the
> ambient
>  > > temperature in Table 310.16. We also have the third adjustment of Table
>  > > 310.15(B)(2)(c) in the 2008 NEC for conduit close to rooftops. Even
> if you
>  > > are excused from using the 2008 NEC by a jurisdiction, the 2005 NEC has
>  > > 310.10 FPN2 that generally recommends a 17C adder on ambient
> temperature.
>  > >
>  > > The NEC has not had any explanation as to what ambient temperature
> to use
>  > > until the 2008 NEC in the FPN to Table 310.15(B)(2)(c) when it
> referenced
>  > > ASHRAE data in an incorrect way. To be consistent with the Copper
>  > > Development Industry, we have put a proposal into the 2011 NEC to
> use the
>  > > ASHRAE 2% design temperatures. These values can be downloaded at
>  > > www.copper.org.
>  > >
>  > > Summarizing in an example:
>  > >
>  > > Assume that 8 current carrying conductors, with and Imax of 10 amps [as
>  > > defined by 690.8(a)], are in a conduit in direct sun 4" off the
> roof deck in
>  > > Palm Springs, California. What must be the 30C ampacity of the
> conductor to
>  > > meet the requirement?
>  > >
>  > > Answer:
>  > >
>  > > I(30C) = 10A/(conduit fill adjustment)/(Temp adjustment--direct sunlit
>  > > conduit)
>  > >
>  > > Conduit fill adjustment factor = 0.7 (70%)
>  > >
>  > > Direct sunlit conduit temperature = +17C above ambient
>  > > 2% Design Temp for Palm Springs = 44.1C (ASHRAE 2005 Fundamentals)
>  > > Design temp = 44.1 + 17 = 61.1C --corresponds to a 0.58 factor for 90C
>  > > conductors
>  > >
>  > > I(30C) = 10A/0.7/0.58 = 24.63 amps -- minimum conductor size is 14 AWG
>  > > (barely)
>  > >
>  > > Most inspectors will quickly cite the fact that the ampacity cannot be
>  > > greater than the 75C column, so we check to make sure (nearly
> always is just
>  > > fine). The 75C column says that 14 AWG wire can handle 20 amps (Imax is
>  > > 10amps) at 30C but the asterisk limits our overcurrent protection
> to 15 amps
>  > > (since the module has a 15 amp max fuse rating, we are already
> using the
>  > > required 15 amp device).
>  > >
>  > > The upshot is that even a 14 AWG 90C conductor works in almost the
> hottest
>  > > climate in the U.S. as long as only 8 conductors or less in
> conduit, conduit
>  > > is at least 4" above roof, and no more than 10 amps flowing through
> it. Most
>  > > contractors will use 10AWG for small systems and occasionally
> 12AWG. 10AWG
>  > > makes it simple since it meets all wiring options in today's smaller
>  > > systems. 12AWG works in many cases, and, as our example shows, even
> 14 AWG
>  > > works in some circumstances (we're talking ampacity, not voltage
>  > > drop--that's a different issue). In large systems, generally we
> specify the
>  > > minimum wire since it adds up after a few miles of conductor.
>  > >
>  > > Now wasn't that fun--I can't believe anyone could be put to sleep
> by that
>  > > (maybe want to commit suicide, but no sleeping here). The short
> answer is
>  > > that it is complicate and not well organized in the code because the
>  > > majority of wiring systems are indoor. PV and HVAC systems are the
> two most
>  > > common outdoor wiring systems requiring these calculations. Most
> electrical
>  > > engineers doing HVAC wiring are just now learning this stuff.
>  > >
>  > > Most inspectors will quickly cite the fact that the ampacity cannot be
>  > > greater than the 75C column, so we check to make sure (nearly
> always is just
>  > > fine).
>  > >
>  > > --- other stuff cut ---
>
>
>
> _______________________________________________
> List sponsored by Home Power magazine
>
> List Address: RE-wrenches at lists.re-wrenches.org
>
> Options&  settings:
> http://lists.re-wrenches.org/options.cgi/re-wrenches-re-wrenches.org
>
> List-Archive: http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org
>
> List rules&  etiquette:
> www.re-wrenches.org/etiquette.htm
>
> Check out participant bios:
> www.members.re-wrenches.org
>