Meters vs. Displays - Was: Batteryless inverter recommendation [RE-wrenches]

[Joel Davidson]

<x-flowed>

Hello Matt,

I don't know what you mean by cyclometer style.

The GE meter is model number 1-70-S 2.
Cat No 721X070088
91 530 279
Parker Metering Specialists
Lodi, CA
Tested 6-07

Why is it unbelievable that the inverter display matches the generation 
meter and PV Watts being so close?

PVWATTS uses algorithms PV engineers have been using for decades. If you put 
in the right derating, azimuth, and tilt, you get pretty accurate monthly 
and annual results. Here are the actual meter readings:
Date, SunnyBoy, GE
7/23/07, 48.4, -
8/21/07, 270.7, 80
9/20/07, 496.8, 304
10/22/07, 698.2, 504
11/20/07, 820.6, 626
12/21/07, 949.1, 755
1/23/08, 1095, 898
2/22/08, 1237, 1040

Best regards,
Joel Davidson

----- Original Message ----- 
From: "Matt Lafferty"
To: <RE-wrenches at topica.com>
Sent: Friday, February 29, 2008 7:45 PM
Subject: RE: Meters vs. Displays - Was: Batteryless inverter recommendation 
[RE-wrenches]


>
> Wrenches,
>
> Great thread. There have been quite a few really good threads lately. Love
> to see the participation and sharing of knowledge. I'll keep my post 
> limited
> to the topic of "inverter display" data vs. "generation meter" data.
>
> The short answer to Jeff C's question about "any new info since May
> '06?".... Yes. I've learned a &@!* load about different metering
> methodologies and, perhaps more important than anything... Just how screwy
> it is that, in this day and age, there are still huge gaping holes in the
> standards for solid-state meters. Yes, this includes "Utility Revenue 
> Grade"
> models. I'll leave the muddy details out of this conversation. There is
> ongoing work on that front by utility and related folks. (This affects
> waaaaaay more than RE. In fact, RE is not even a flea on the back of this
> issue...)
>
> There is also an in-progress development of testing requirements for
> inverter integrated +/- 5% Class meters. For the foreseeable future, I
> recommend installing at least a meter socket capable of accepting a
> "standard" socket-type meter in the output circuit of all grid-connected 
> PV
> systems. Whether or not a generation meter is required by the utility or
> incentive program. Whether or not you actually install a meter. At least 
> put
> the meter socket in there. It's cheap. I care less and less about whether 
> or
> not people take my advice with each passing day, so... Take that advice or
> leave it.
>
> Another thing I have learned is how many brand new meters flunk utility
> meter shop testing and are returned to the manufacturer for replacement
> instead of being deployed. And those are self-certified and calibrated by
> the mfrs. Some utilities physically test every single meter and some only
> test sample quanitities. If the sample fails, then the batch fails. Along
> the same line is something else I've learned, which explains a lot of
> "unanswered questions"... The meters you get from Austin or Hialeah or 
> other
> such supplier, may or may not be fully tested and calibrated to the same
> accuracy standard. In a nutshell, this means that apparently identical
> meters installed in paralell on the same circuit, may or may not record 
> the
> same kWH readings over time. The differences might or might not be
> significant, which in and of itself is subject to ones definition of
> significant. The important thing to insist on is a certificate stating 
> what
> standard the meter you get is tested to. For "standard" socket type 
> meters,
> you're looking for ANSI C12.20 and an accuracy class.
>
> And, oh by the way, if you are installing one of the 3rd Party monitoring
> systems that has an independent meter and they "calibrate" the meter data
> feed to match the inverter display... Now you really don't know what the
> truth is... But hey, this is solar after all, so maybe some don't care to
> know the truth. Or maybe it doesn't matter in the big picture that I can't
> see because, well, I care too much about the truth. Speaking of truth, I
> don't think I've seen any module mfr reps on the list in like forever... 
> But
> back to metering.
>
> Joel D's SB2100 had approximately a 0.57% difference in kWH between the
> inverter and the generation meter. (740.2 vs. 736 respectively.) This is
> over a 5 month period. Assuming Joel's generation meter is fully certified
> and calibrated, this is a reasonable discrepancy in my opinion.  Nightime
> and standby losses being accounted for on the generation meter but not on
> the inverter display could account for the entire difference. In a 
> nutshell,
> they may not actually be measuring exactly the same things.... Either way,
> at least for that 5-month period, at Joel's place, the SB2100 kWH display
> was nearly identical to his GE I70. (Joel, do you happen to have the
> "cyclometer" style? According to the Maryland Public Service Commission, 
> the
> cyclometer style I70 is not recommended for some unstated reason... Maybe
> they found out that is is calibrated to SMA SB2100 inverters!  LOL) Joel, 
> I
> hope you feel special. Nobody else's inverter displays match a 
> socket-based
> meter that closely over a 5-month period... I don't know which part of the
> story I find more unbelievable... The part about the inverter display
> matching the generation meter or the part about PV Watts being so close!
> LOL
>
> Brad B & Ian W report "wildly" different and greater inverter display
> readings than the utility generation meters.  Both of these inverters are
> reported to be SMA SB700's. (120 VAC inverters with three output power 
> level
> ratings... 460, 600 & 700 Watts).  I suspect a couple of things here.
>
> Brad & Ian: Are you comparing readings for the same time periods? You 
> didn't
> happen to turn the system on before the utility got there with a 
> generation
> meter, did you???  I know how we all hate waiting for them.... Brother 
> Brad
> did say it was "his system". This item was #1 on my original post from 
> 2006
> for a reason.  :-)
>
> Another issue is that a few kWH difference over a short period of time on 
> a
> relatively small system will equate to a large percentage. Comparing 
> values
> for identical periods over longer time will provide a more reliable basis
> than just a few weeks. Also related to this is whether or not the 
> generation
> meters are "bi-directional". In winter months, it is common for inverters 
> to
> spend a greater part of their time in "standby", which can result in 
> higher
> percentage tare losses, thereby subtracting from the total generation.
>
> Assuming the measured time periods were identical, the reported "wild"
> difference smells like a software or configuration issue in the inverter
> meter/display. In looking at the difference between the lowest & highest
> output power ratings, I suspect the display is showing a value that is
> calculated based on a configuration that is different from actual.  (For
> instance: If the default setup is for a 700 Watt configuration, but you've
> actually installed a 460 Watt, it could be possible that the value on the
> display is calculated as if the system were larger than it is... Up to 52%
> larger.) John B???
>
> I won't mention names here, but there once was an inverter manufacturer 
> who
> shipped a few inverters with brains from other inverters... Which 
> basically
> ended up resulting in some wild differences between "expected" and
> "measured" energy & power... Something with the magnitude of 75:125 or
> 100:225.... Don't remember the specifics.... Just the fact that the 
> systems
> actually ran, the installers didn't catch it... The 3rd party monitoring
> company setup & calibration procedure resulted in data feeds that matched
> the inverters' displays... It was a classic example of a mfr slip-up
> followed by inadequate trained installers being directed to follow a 
> flawed
> 3rd party procedure which caused a whole bunch of head-scratching, 
> resorting
> of the books, embarrassment all the way around, and wasted time by all. 
> "So,
> explain to me again why this 80 kW PV system is putting out 150 kW at 3 PM
> in November..." Ooooopsie! The sick part is that certain folks thought it
> just meant the systems were of "extra high quality". I'm sure we all
> understand that these folks are the bosses... Advanced degree types... Not
> Wrenches.... Like I said... I'm not gonna mention any names...
>
> Sunshine wishes to all....
>
> From the Hammock,
> Matt Lafferty
>
>
> -----Original Message-----
> From: Jeff Clearwater, Village Power Design
> Sent: Tuesday, February 26, 2008 5:31 PM
> To: RE-wrenches at topica.com
> Subject: RE: Batteryless inverter recommendation [RE-wrenches]
>
>
> Hi Brad,
>
> At 1:20 AM +0000 2/27/08, Brad Bassett wrote:
>>I now have a few weeks of time on my SB700 system that also has a
>>utility owned digital revenue meter on it. So far the SB meter reads
>>145% of the revenue meter reading.
>
>
> I've brought this to the attention of the List in the past (See 5/18/
> - 5/20/06 - KWH meter discrepancy).  Matt replied with a very complete 
> list
> of possible causes (below).
>
> We do really need to get to the bottom of it.  Any more thoughts or 
> updates
> since May of '06 Matt?  Others?
>
> Best,
>
> Jeff Clearwater
> Village Power Design
>
> At 10:04 AM -0700 5/20/06, Matt Lafferty wrote:
>>Wrenches,
>>
>>I see a couple of things "missing" from the descriptions in this thread
>>for a conclusive determination of the true cause for the discrepancies...
>>
>>1) Hours of operation.  In many cases, even though it is generally
>>"prohibited", the PV systems will have a number of hours or days
>>"head-start" on the PV Generation Meters.  I've seen numerous cases
>>where this accounts for over a month of production on the inverters by
>>the time the PV Meter is installed.  If your comparisons have been made
>>on relatively new systems, there is a good chance some of your
>>discrepancy is related to this.  It is important to take a reading on
>>both meters at the same time, then take another reading some time later
>>for proper comparisons.  The longer the time interval between readings,
>>the more accurate your comparison results will be.  (I'm not saying
>>they will be identical by any means.)
>>
>>2) Most revenue grade kWH meters account for power factor losses to
>>some degree.  What this means is that a site with a low power factor
>>actually "uses" more energy to do the same "work" as far as the meter is
> concerned.
>>This is a double-bite for our purposes... The Customer is paying for
>>kWH that they wouldn't need if the PF was 1.00 and their PV Meter is
>>recording the opposite effect. This occurs regardless of whether the
>>low power factor is "incoming" or "site-caused. I can't speak for the
>>Inverter Manus, but my observations, testing, and comparisons have led
>>me to the conclusion that their metering does not account for this as
>>accurately as revenue grade kWH meters.  (Note: It is common for
>>utilities to deny this and/or claim that the low PF is caused by the
>>Customer.  I've seen them claim that they are "losing revenue" because
>>of this issue.)
>>
>>3) Most revenue grade kWH meters are very accurate in the "forward"
>>direction (+/- .5%), but wildly vary in their accuracy when running
>>"backward". This aspect is huge on a number of fronts and there isn't
>>room to describe the plethera of scenarios here. You may be wondering
>>"why would my PV meter run backward?"... Simple.  At nite, the inverter
>>uses some AC power... A very small amount based on the Manus, but some
>>nonetheless.  It is impossible to tell here what that effect has on the
>>"dial readings".  The Inverter displays do not count this.  For
>>starters, you will have to know whether or not the PV meter is detented
>>or non-detented.  After that you will have to find out whether or not
>>it is calibrated to meter "bi-directional" or not and the accuracy
>>tolerances for both.  Good luck getting that info!
>>
>>4) If you have a 120V Inverter, you need a 120V meter to properly meter 
>>it.
>>Running the noodle through one leg of a 240V meter will not provide
>>accurate results.  Period.
>>
>>5) Electro-mechanical vs. "calculated".  This is another spot where the
>>differences get wild. Your PV Meter is most likely EM.  Inverter
>>displays/meters are calculated.  The circuitry & programming in the
>>Inverters uses "constants" in order to "do the math".  Contact your
>>Inverter Manu for specifics related to their methodology.  Be sure to
>>let us all know what you find out!  (Hint: This is very closely related
>>to #2 above.)
>>
>>Here are the high-points of my testing and research over the years:
>>
>>After eliminating #1 & #4 above, the largest discrepancy factor is a
>>combination of #2 & #5. Once you've isolated and quantified them,
>>multiply the Inverter meter reading by the average Power Factor.  You
>>will be amazed how closely the result comes to matching the PV Meter
> reading.
>>
>>How to go about doing it:
>>
>>In order to track this, you'll have to get your hands on a recording
>>analyzer that does power factor measurements.  (I.E. Fluke 43 or 
>>similar...
>>~$2K) If you don't have one, and most of us don't, it is possible to
>>request a "power quality" or "meter accuracy" test from your utility
> company.
>>
>>Generally they all do this at no charge.  You will have to tell them
>>specifically what aspects you want results for and the interval that
>>measurements are taken... Try to talk directly to the Power Quality
>>Tech and let him/her know what you are trying to do.  They are
>>generally pretty laid back and quite helpful if you aren't pushy or
> "blaming the big bad utility".
>>They will let you know what they can measure/record at what intervals
>>and for how long.  Generally speaking, the more channels you want
>>measured reduces the length of time and increases the interval period for
> your test.
>>After you get a decent relationship going, be sure to try to get the 
>>"data"
>>as well as the standard "report".  Ask for it in a CSV or spreadsheet
>>format.  This may or may not be possible due to "policy".  Generally
>>speaking, you will get a lot more bees with honey, so be sure to have
>>jelly donuts on hand.  Your attitude will play a large part in
>>determining how all this goes. If you know you can't be civil with them,
> send somebody who can!
>>(Hint: "Awww shucks" and "Golly gee" and "I didn't know that... Wow!
>>Thanks!" will get you a long ways toward your goal... These guys are
>>"tech-heads" by nature and they can talk way over your head in most
> cases...
>>Way over mine, anyway. The "Awww shucks" thing plays nicely in getting
>>the conversation onto a "plain English" level.)
>>
>>Some utilities have power quality recorders available to Electrical
>>Contractors on a loaner basis for use in their service territories.
>>Check into this option.
>>
>>OK. Off to spend the rest of the weekend campin at a Hot Springs.  Have
>>a good one, ya'll!
>>
>>Pray for Sun!  (I'd prefer not to have my sleeping bag soaked)
>>
>>Matt Lafferty
>>Universal Energies Institute
>>mlafferty at universalenergies.com
>>(415) 252-0343 Phone
>>(916) 914-2247 Fax
>>
>
>
>
>
>>I now have a few weeks of time on my SB700 system that also has a
>>utility owned digital revenue meter on it. So far the SB meter reads
>>145% of the revenue meter reading.
>>
>>Brad
>>AEE Solar
>>
>>
>>Joel Davidson wrote:
>>
>>>
>>>  I have 20 Siemens SP75 modules feeding into an SB2100U inverter and
>>> a GE  I70  kWh meter. From September 20, 2007 to February 22, 2008
>>> the SB2100U  measured
>>>  740.2 kWh and the GE I70 measured 736 kWh. (PVWATTS estimates 748
>>> kWh at  81%  DC-to-AC derate.) Looks good to me.


- - - -
Hosted by Home Power magazine

To send a message: RE-wrenches at topica.com

Archive of previous messages: http://lists.topica.com/lists/RE-wrenches/read

List rules & how to change your email address: www.mrsharkey.com/wrenches/etiquette.php

Check out participant bios: www.mrsharkey.com/wrenches/

Moderator: michael.welch at homepower.com
--^----------------------------------------------------------------
This email was sent to: michael.welch at re-wrenches.org

EASY UNSUBSCRIBE click here: http://topica.com/u/?bz8Qcs.bz9JC9.bWljaGFl
Or send an email to: RE-wrenches-unsubscribe at topica.com

For Topica's complete suite of email marketing solutions visit:
http://www.topica.com/?p=TEXFOOTER
--^----------------------------------------------------------------


</x-flowed>