Announcement

Collapse
No announcement yet.

Pyranometer with modbus connection

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Pyranometer with modbus connection

    Hi,

    I was interested in collecting some data from a pyranometer, and wondering if anyone had any experience with any pyranometers that are solid. We like Modbus in general so that would be the preferred method but interested in any recommendations for quality instruments.

    Thanks, John

  • #2
    Depends on price. For quality, factory support and durability, Eppley is about as good as it gets. But you better have deep pockets.

    Lots of different types and mfgs. depending on application. I used and maintained an Eppley PSP and an Eppley "Black & White ands a

    Kipp & Zonnen are also quality instruments with a somewhat lower price.

    For a lot less $$ and almost as much accuracy and precision, but probably more frequent calibration, there are a lot of silicon photodiode instruments. About the most frequently found in the field are from Li-Cor. A decent reputation and a favorite due to it's relatively (compared to Eppley) low price and available support.

    Lots of folks, including a few around here use a Davis Pro II plus weather station. For a relatively low admission price ( ~~ $800 on up, depending on bells/whistles/software) you'll get a decent weather station that's easy to use and record/store data that's about as accurate as warranted, including a pyranometer that's reasonably accurate. Just plan on replacing the pyranometer every couple of years for ~ $125-150 or so and calibrate it against the old one.

    A word on pyranometer accuracy: Even the Eppley's will be not much better than +/- a % or 2, (And they'll tell you 5% officially, in writing) and even then that's only with frequent (1X/yr. or so) recalibration. That's not a knock, just reality of the task.

    The silicon devices are almost as good (maybe +/- 2% or better if you're lucky, but much less precision as f(cloud/atmospheric conditions), for a lot less $$, and have what can be considered the advantage of being mostly insensitive to off horizontal orientation and operation.

    FWIW, if you have a PV array, or even a single panel and a way to get a reasonably accurate estimate of that array's (panel's) cell temp., measuring the array's current can give a pretty accurate estimate of irradiance under a clear sky. I've got one of the Davis instruments next to my array. After corrections for array temp. and pyranometer temp. the array's current under clear skies is just about directly proportional to what the Davis pyronameter reads to a degree I found surprising.

    Comment


    • #3
      Wow J.P.M that's a lot of experience you have with these - thanks! I have heard the midrange value level is the Davis before. When you say the silicon devices, these are the ones that do not have a clear plastic dome, or is this the component inside the device that is silicon?

      Good to know about the DC current measurement too on the panel, interesting. What I have been working on is a neural network module for our monitoring platform that can zero in on performance based on weather and just watt-hours in a given time frame. But I have not added the pyronometer input for that learning, so am interested in testing that.

      Comment


      • #4
        Originally posted by jwrgorman View Post
        Wow J.P.M that's a lot of experience you have with these - thanks! I have heard the midrange value level is the Davis before. When you say the silicon devices, these are the ones that do not have a clear plastic dome, or is this the component inside the device that is silicon?

        Good to know about the DC current measurement too on the panel, interesting. What I have been working on is a neural network module for our monitoring platform that can zero in on performance based on weather and just watt-hours in a given time frame. But I have not added the pyronometer input for that learning, so am interested in testing that.
        FWIW, you're welcome. Short ans. to your questions, at least as the answers may relate to the solar energy application of pyranometers: Common silicon devices are smaller and usually not under a dome. Perhaps coincidentally, the smaller devices cost less. Davis is close to the bottom on price.

        On array (panel) current as a measure of irradiance, two things: Cell temp. is not ambient temp. and, any estimates/correlations using that method including temp. corrections, will be of est. P.O.A. (Plane Of Array) irradiance which, while it may be bottom line what you're looking for, will not be the same as most published irradiance data which is usually reported as G.H.I (Global Horizontal Irradiance) with the user then transposing/converting the GHI to POA for the orientation of the array or device using any one of available algorithms or models. See NREL for more info.

        I suppose from a technically correct standpoint, the silicon diode devices like the Davis and Li-Cor and simpler stuff are not considered true pyranometers because of the definitional requirement of true pyranometers requiring a response to the incoming solar radiation that is independent of the wavelength of the incoming solar radiation and also be independent of the angle of incidence of all radiation for beam, diffuse and any albedo sources. The silicon devices' response are sort of representative but not truly wavelength independent. See a response curve for the graphic. Most folks don't make that distinction, or maybe need to.

        Perhaps interestingly, the silicon devices may actually be closer to the response of a PV array to irradiance than a true pyranometer, but not give as true a picture of the value of GHI or POA irradiance as the thermopile devices. That probably doesn't matter a whole lot for day/day backyard stuff but good to keep in mind if things require more detail and accuracy.

        Best suggestion I can give for answers to your questions is to first see the first few sections of "Solar Engineering of Thermal Processes", by Duffie & Beckman, ISBN # 0-471-51056-4. The info there may be a bit outdated with respect to the more recent silicon devices, but still good background info. Then, armed with that background, see the product literature from outfits I mentioned which, while not bad, is from those with skin in the game and stuff to sell. Caveat Emptor.
        Last edited by J.P.M.; 01-09-2018, 04:00 PM.

        Comment


        • #5
          Hi J.P.M., that makes a lot of sense. I am all for a lower cost option, especially when the physical device is similar in function to the device creating the electricity you're benchmarking - but making sure we're dealing with apples is important. Just so I get this right - the P.O.A. irradiance - is that often taken from a unit mounted in the same plane as the array OR is it calculated from a horizontal mounted unit? I can see both datapoints being interesting. Also the thermopile devices that are more frequency independent, those are always horizontally mounted?

          Comment


          • #6
            Originally posted by jwrgorman View Post
            Hi J.P.M., that makes a lot of sense. I am all for a lower cost option, especially when the physical device is similar in function to the device creating the electricity you're benchmarking - but making sure we're dealing with apples is important. Just so I get this right - the P.O.A. irradiance - is that often taken from a unit mounted in the same plane as the array OR is it calculated from a horizontal mounted unit? I can see both datapoints being interesting. Also the thermopile devices that are more frequency independent, those are always horizontally mounted?
            Short answers: Purists use horizontal orientation for resource assessment needs like determining clearness indices and such like. They convert the GHI to POA for determining efficiency and other needs in solar energy generation work/estimating. Thermopile type devices are always horizontally oriented.
            .
            P.O.A., Plane Of Array irradiance, is a necessary quantity when dealing with arrays or other flat plate solar devices and their efficiency. GHI, Global Horizontal Irradiance is a necessary quantity for investigating things solar resource availability.

            Historically, because most solar resource tools - that is pyranometers - were once mostly of the thermopile type that, because of their design, and the idea that natural convection currents inside the instrument (under the dome) are different in different orientations and so affect the devices sensitivity and accuracy as well, and also due to the time honored convention of reporting GHI, the standard has developed of using all pyranometers a horizontal orientation. But, the rub with that was/is that because most solar collectors were not horizontal, a very large body of knowledge developed around and dealing with converting GHI to POA irradiance. About the best concise but reasonably complete treatment of the basics of all that is in Duffie & Beckman. See NREL for additional info.

            Now, to your questions:

            I suppose POA day long insolation as gathered by silicon devices mounted on an array will probably be sufficient, but will probably be different than day long total insolation that starts with daylong GHI from the same device in a horizontal orientation (including the Davis weather station pyranometers), and then converted that GHI to POA via readily available models and algorithms. The cosine response of the device in the plane of the array may well be slightly different and albedo certainly will be different, maybe more, maybe less, depending on array location and orientation. There are other things that will affect the readings horizontal to POA as well. Purist types I know, and most all of the literature use GHI as the basis from any and all types of pyranometer - silicon and thermopile type - and convert to POA as needed. The amount of software available to do the conversion modeling has small variations if the models are comprehensive and the weather data used is reasonably accurate, particularly with respect to precipitable water vapor and atmospheric turbidity.

            Starting with GHI and converting just keeps everyone on the same page, dealing with apples from the same orchard and working to mostly the same assumptions. After a while, you'll get to appreciate the advantages.

            Comment


            • #7
              Thanks J.P.M. - didn't think of convection currents inside the instrument, that's a lot fine detail. OK will stick with GHI for now, I am reviewing here as well : https://pvpmc.sandia.gov/modeling-st...oa-irradiance/

              Comment


              • #8
                Originally posted by jwrgorman View Post
                Thanks J.P.M. - didn't think of convection currents inside the instrument, that's a lot fine detail. OK will stick with GHI for now, I am reviewing here as well : https://pvpmc.sandia.gov/modeling-st...oa-irradiance/
                The Sandia reference you include is a good start, but there's a lot of detail not shown that will be needed to get the most out of it. I don't want to sound like a broken record, but Duffie & Beckman is still the best place to find most of what you'll need in one place.

                Using GHI and converting to POA will be more work up front, but the extra detail may be worth the effort if you need more than POA at one orientation - for comparison at different orientations for example. D & B have a good discussion of that.

                Comment


                • #9
                  Thanks J.P.M. I hear you will be checking out Duffie and Beckman!

                  Comment


                  • #10
                    Originally posted by jwrgorman View Post
                    Thanks J.P.M. I hear you will be checking out Duffie and Beckman!
                    Learn the basics. Set goals. Do you want to learn about alternate energy or how to monitor it ? The two are not the same.

                    Comment


                    • #11
                      I am starting with a Rika RK200-03 unit, which I believe is a thermopile device - it has Modbus and you can configure a SolarNode unit (in this case running on a Raspberry Pi) to acquire data from it at whatever schedule you like. This I think will give a baseline, and we should be able to compare empirical PV performance against it. But I'm interested in testing the silicon-based devices as well.
                      Attached Files

                      Comment


                      • #12
                        Originally posted by jwrgorman View Post
                        I am starting with a Rika RK200-03 unit, which I believe is a thermopile device - it has Modbus and you can configure a SolarNode unit (in this case running on a Raspberry Pi) to acquire data from it at whatever schedule you like. This I think will give a baseline, and we should be able to compare empirical PV performance against it. But I'm interested in testing the silicon-based devices as well.
                        I've no experience with Rika, but that doesn't make it bad. How much cost ? Speed of warranty service if needed ? Don't forget temp. compensation and 2 yr. recalibration. Also, suggest if you get a silicon device for comparison, consider Li-Cor as they are widely used.

                        On PV performance, any ideas on what parameters you plan to measure and how ? FWIW, suggest reviewing the journal "Solar Energy" for ideas and review of some testing/measurement methods.

                        Comment


                        • #13
                          I do not know cost yet - was purchased for me - but I think it was pretty reasonable compared to other thermopile units. It feels very heavy and solid, and is one of many industrial sensors in their gloassy catalogue - but we'll see how it goes. I hear you about warranty and maintenance costs - they get real if you're not dealing with quality, durable gear that can sit outside 24/7/365.

                          As far as inputs, we're seeing a lot of good correlation of performance with straight weather and the kilowatt hours produced by the array. I think adding irradiance (it is insolation when it comes to watt-hours per meter squared right?) as an input is only going to help. the software module in development called SolarQuant uses a neural network to weight the inputs so it adapts to what it seems over time. This is an early screenshot after admittedly NOT a lot of testing... but the standard deviation is respectable between forecast and actual. we'll use coefficient of variation going forward to keep units apples-to-apples.
                          Attached Files

                          Comment


                          • #14
                            Originally posted by jwrgorman View Post
                            I do not know cost yet - was purchased for me - but I think it was pretty reasonable compared to other thermopile units. It feels very heavy and solid, and is one of many industrial sensors in their gloassy catalogue - but we'll see how it goes. I hear you about warranty and maintenance costs - they get real if you're not dealing with quality, durable gear that can sit outside 24/7/365.

                            As far as inputs, we're seeing a lot of good correlation of performance with straight weather and the kilowatt hours produced by the array. I think adding irradiance (it is insolation when it comes to watt-hours per meter squared right?) as an input is only going to help. the software module in development called SolarQuant uses a neural network to weight the inputs so it adapts to what it seems over time. This is an early screenshot after admittedly NOT a lot of testing... but the standard deviation is respectable between forecast and actual. we'll use coefficient of variation going forward to keep units apples-to-apples.
                            Understood, I think. What you have is probably a lot less expensive than an Eppley SPP. Jut take care of it. Part of that means set it up right and then don't mess with it except for cleaning, maint. and calibration. Based on experience, mounting it in such a way that it gets irradiance in the same way and quantity as any solar array performance you're analyzing, and in such a way that access is easy will help, particularly if cleaning the cover is a priority - which it will be if you want good data.

                            I don't understand what you mean by good correlation with straight weather, or even what "good" means in that context. If you are looking to measure array performance as f(weather), including irradiance, you will need, as a minimum, temp. and wind vector info gathered as close to the array as possible, and also some estimate of the P.O.A. irradiance at the array - and that means data gathered right next to the array, not, say, 100 m away on a roof somewhere for example.

                            You will also need some way to convert GHI irradiance from the pyranometer to POA irradiance. Using the pyranometer GHI without conversion to POA is useless and will lead to inaccurate results that will really screw you up. See Duffie & Beckman, chap. 2 for details and models to convert GHI to POA. My apologies if you already know that.

                            I've found the HDKR model that you'll find described in D & B to give results that seem to make sense. Also, my clear sky GHI measurements seem to agree quite well with the Bird model/estimates for clear sky GHI. You'll find the bird model as an excel spreadsheet from the NREL website. (Richard E. Bird, Clear Sky Broadband Solar Radiation Model).

                            My apologies if I'm missing something else here, but I'm also a bit confused by your statement that adding irradiance as an input will help. My confusion come from the fact that if your software module is measuring efficiency of some sort or in some fashion, not only is irradiance data helpful, it's essential - you can't get there from here without it (and also not without some way to convert GHI to POA irradiance). Since efficiency = output/input, and since input is GHI converted to --->>> POA, and GHI is measured by the pyranometer, seems kind of hard to get a model to work without GHI data from a pyranometer. Note: Clear sky data from models will always be an estimate. Modeled data can usually be close but it'll almost never match real data, often by a few %. In spite of what they may look like and in spite of what you may think, all clear skies are different. Trust me on that one.

                            You will also need some way to either measure or estimate cell temperatures, either integrated over the array or some average value. FWIW, I've found that the Sandia model as described in
                            " Photovoltaic Array Performance Model " by D.L. King, et al., Sandia National labs. does a pretty fair job in a reasonable way of estimating array temps. that I've been able to verify. There is a lso a way to verify or estimate cell temps. from panel voltages. The biggest uncertainly in cell temps is the error or uncertainty in the effect of the wind vector's variability and uncertainty and it's effect on heat transfer from the array and hence the array's temp profile. After several hundred measurement events over the last 4 years, I've taken/crosschecked/correlated enough data to convince myself of the worth and validity of that (Sandia) model using data taken with the Davis Instruments weather station located about 4 ft. north of my array's N-S centerline at about the same elevation as the north (highest) edge of the array.

                            Good luck.
                            Last edited by J.P.M.; 02-14-2018, 09:36 PM.

                            Comment


                            • #15
                              I am listening carefully J.P.M. thank you! I really like that quote also: "All clear skies are different" - there is a lot of nuance to this science I guess. The POA sensor may be more appropriate for what we are trying to do now that you mention the conversion from GHI to POA. This is what we are trying to do and out assumptions: when you look at solar generation over a day for any given site, there are many computable reasons as to why you got so many kilowatt hours out of the system. The time of the year, the time of the day, the sky conditions, the temperature of the air, the temperature of the panels, the wind direction, the shading from a tree 60m away to the southwest, the kinds of clouds, the pitch and orientation of the array, the brand of module, the brand of inverter, the humidity, the barometer readings, the way it is wired, etc.etc. Let's call these the "context" of the energy production. So while some of these can give you very exact reasons (lower/higher conductivity of the module cells, angle of the sun to POA) why you would get a certain number of watt-hours, there are probably many that add a lot of variance to that figure. And because some elements of the context depend on other elements of the context, it's complex. For the average solar PV user (I have a bog-standard 1kW SMA Sunny Boy flushmount multi-crystalline system for example) I don't think they have too many easy ways to judge all these things or get a simple forecast of what tomorrow looks like in terms of watt-hours. Yes there are becoming available. Certainly the NREL labs and NIST and other science foundations are going to have more sophisticated models. So our goal is not to do that, but provide a "point and shoot" version of energy data analysis that uses the newly available neural networking capability (that is all over the news these days) and apply it to this basic challenge. The theory is that by presenting a neural network with a set of inputs that describe the "context" as well as the answer of how many watt-hours per produced, it will naturally weight those inputs in ways to zero in on that answer. Then, you can approximate a "context" for tomorrow let's say and feed that to the trained network- and it should come up with some estimates. Clearly it will need iteration and possibly 12 months of data and maybe years of experience to get it really right, but early testing on this is looking promising. my question to you as an expert on this science is: do you think there is any utility for the average person to get some decent numbers, some reasonable idea of what tomorrow or this week looks like in terms of generation from their specific array? For example - you mention wind direction and module temperature right at the site, not remote by 100m, is that part of the "context" going to be critical in getting closer to our goal? If so we need to put a thermistor on our test panels, and look at integrating a device like the Davis weather stations that we can locate near the array. Also: do you recommend NOT using GHI thermopile in this case but going directly to a POA device, possibly silicon based device, mounted in the plane of the array? Thanks, John

                              Comment

                              Working...
                              X