CLOSED: [2024-01-07 Sun 18:23] :PROPERTIES: :ID: 2024-01-07-pv-2023 :CREATED: [2024-01-07 Sun 17:15] :END: :LOGBOOK: - State "DONE" from "TODO" [2024-01-07 Sun 18:23] :END: In the [[id:2023-08-10-our-new-PV][previous article on our PV]], I've summarized many things related to the planning and construction phase and ended with the figures from the first six months. I assume, you went through this article before you continue below. Here, I want to deliver the figures of the year 2023 which is more or less equal to the total run-time of our PV so far. **** Diagrams from Victron VRM :PROPERTIES: :END: Here are the visualizations from the web interface of Victron (VRM): #+CAPTION: Victron VRM: System overview 2023 #+ATTR_HTML: :alt Screenshot from Victron VRM with bar charts of various values. #+ATTR_HTML: :align center :width 630 :linked-image-width 1000 [[tsfile:2024-01-07T17.09.02 1 System overview -- screenshots publicvoit statistics bwg.png][2024-01-07T17.09.02 1 System overview -- screenshots publicvoit statistics bwg.png]] The maximum for the solar is in June and July which peaks at 1660 kWh per month. Following the blue line teaches us that our battery was mostly full in summer and got down to an average of seventy percent in winter and autumn. #+CAPTION: Victron VRM: Consumption 2023 #+ATTR_HTML: :alt Screenshot from Victron VRM with bar charts of various values. #+ATTR_HTML: :align center :width 630 :linked-image-width 1000 [[tsfile:2024-01-07T17.09.35 2 Consumption -- screenshots publicvoit statistics bwg.png][2024-01-07T17.09.35 2 Consumption -- screenshots publicvoit statistics bwg.png]] You see three maximums of consumption: 734 kWh in Jannuary for heating, 684 kWh in July for cooling and 709 kWh in December for heating. Battery was very important from March to October. This was also the period where our house was more or less independent from the grid - except for one single longer period of rainy days in early August 2023. #+CAPTION: Victron VRM: Solar 2023 #+ATTR_HTML: :alt Screenshot from Victron VRM with bar charts of various values. #+ATTR_HTML: :align center :width 630 :linked-image-width 1000 [[tsfile:2024-01-07T17.09.54 3 Solar -- screenshots publicvoit statistics bwg.png][2024-01-07T17.09.54 3 Solar -- screenshots publicvoit statistics bwg.png]] We delivered power to the grid from March to September - only a small amount in October. In dhe diagram above, you can clearly see the amount of power from the battery in blue. It has a high impact from February until November with exception of the sun-intense months in summer. So this is, where you profit from a battery: February→April and September→November. Of course, you always get advantages when you want to be as independent as possible from the grid (at night) and in case there are power outages. At least to my knowledge, there was no outage here in 2023. But I don't have any workflow set up where I would be alarmed. And for the night, you clearly see it in the numbers: without the battery, we would have used much more power from the grid. #+CAPTION: Victron VRM: Grid 2023 #+ATTR_HTML: :alt Screenshot from Victron VRM with bar charts of various values. #+ATTR_HTML: :align center :width 630 :linked-image-width 1000 [[tsfile:2024-01-07T17.10.15 5 Grid -- screenshots publicvoit statistics bwg.png][2024-01-07T17.10.15 5 Grid -- screenshots publicvoit statistics bwg.png]] The grid comsumption is a good overview on the level of independence from the grid. The small amount of grid energy from April to September is most likely caused by used energy by the power inverters in order to keep in-sync with the grid. If you have more details on that, please drop me a line. **** Figures Here are the monthly figures I wrote down: #+NAME: blog-ESS-Stats-2023 | | Month | →Grid | Grid→ | Solar | Consumption | Grid→Con | Batt→Con | Solar→Con | Sol→Grid | Sol→Batt | max batt %/day | max kWh solar/day | max consumption/day | days <100% batt | | ! | month | togrid | fromgrid | solar | consumption | confromgrid | confrombatt | confromsolar | solartogrid | solartobatt | maxbattloadday | maxkwhsolarday | maxconsumptionday | battnotfulldays | |---+---------+--------+----------+-------+-------------+-------------+-------------+--------------+-------------+-------------+----------------+-------------------+---------------------+-----------------| | | 2023-01 | 3.2 | 576 | 193 | 734 | 547 | 63 | 124 | 2.5 | 66 | 19 | 10.6 | 29.59 | 29 | | | 2023-02 | 9.3 | 293 | 435 | 678 | 274 | 202 | 202 | 8.1 | 224 | 39 | 20.4 | 35.68 | 27 | | | 2023-03 | 8.0 | 82 | 818 | 631 | 75 | 282 | 274 | 185 | 360 | 46 | 40.9 | 24.24 | 8 | | | 2023-04 | 491 | 35 | 1096 | 575 | 32 | 251 | 292 | 489 | 316 | 51 | 60.9 | 25.72 | 7 | | | 2023-05 | 764 | 32 | 1363 | 565 | 29 | 214 | 321 | 761 | 280 | 66 | 76.51 | 34.73 | 5 | | | 2023-06 | 1073 | 27 | 1664 | 550 | 25 | 183 | 343 | 1071 | 250 | 62 | 79.26 | 34.04 | 2 | | | 2023-07 | 945 | 29.7 | 1667 | 684 | 29.7 | 250 | 407 | 943 | 317 | 44 | 76.1 | 32.48 | 2 | | | 2023-08 | 626 | 36.8 | 1311 | 649 | 33.6 | 273 | 342 | 623 | 346 | 55 | 62.2 | 33.74 | 8 | | | 2023-09 | 368 | 32.5 | 954 | 552 | 29.6 | 250 | 273 | 366 | 315 | 53 | 45.3 | 28.78 | 5 | | | 2023-10 | 71.2 | 77.3 | 528 | 492 | 70.0 | 213 | 208 | 69.5 | 250 | 33 | 26.3 | 20.87 | 18 | | | 2023-11 | 5.5 | 363 | 305 | 631 | 344 | 119 | 169 | 4.3 | 132 | 27 | 14.2 | 30.13 | 30 | | | 2023-12 | 1.4 | 623 | 138 | 709 | 580 | 47.3 | 81.1 | 1.1 | 55.6 | 25 | 13.9 | 31.47 | 31 | These are some numbers I calculate from the table above | | Month | Grid→Cons % | Solar→Cons % | Batt→Cons % | Sol/Batt→Cons % | Sol/day | Con/day | Solar/Cons % | %days batt<100% | Days/month | | ! | month | | | | | | | | | numdays | |---+---------+-------------+--------------+-------------+-----------------+---------+---------+--------------+-----------------+------------| | | 2023-01 | 74.5 | 16.9 | 8.6 | 25.5 | 6.2 | 23.7 | 26 | 94 | 31 | | | 2023-02 | 40.4 | 29.8 | 29.8 | 59.6 | 15.5 | 24.2 | 64 | 96 | 28 | | | 2023-03 | 11.9 | 43.4 | 44.7 | 88.1 | 26.4 | 20.4 | 130 | 26 | 31 | | | 2023-04 | 5.6 | 50.8 | 43.7 | 94.5 | 36.5 | 19.2 | 191 | 23 | 30 | | | 2023-05 | 5.1 | 56.8 | 37.9 | 94.7 | 44.0 | 18.2 | 241 | 16 | 31 | | | 2023-06 | 4.5 | 62.4 | 33.3 | 95.7 | 55.5 | 18.3 | 303 | 7 | 30 | | | 2023-07 | 4.3 | 59.5 | 36.5 | 96.0 | 53.8 | 22.1 | 244 | 6 | 31 | | | 2023-08 | 5.2 | 52.7 | 42.1 | 94.8 | 42.3 | 20.9 | 202 | 26 | 31 | | | 2023-09 | 5.4 | 49.5 | 45.3 | 94.8 | 31.8 | 18.4 | 173 | 17 | 30 | | | 2023-10 | 14.2 | 42.3 | 43.3 | 85.6 | 17.0 | 15.9 | 107 | 58 | 31 | | | 2023-11 | 54.5 | 26.8 | 18.9 | 45.7 | 10.2 | 21.0 | 48 | 100 | 30 | | | 2023-12 | 81.8 | 11.4 | 6.7 | 18.1 | 4.5 | 22.9 | 19 | 100 | 31 | |---+---------+-------------+--------------+-------------+-----------------+---------+---------+--------------+-----------------+------------| #+TBLFM: @3$2..@>$2='(identity remote(blog-ESS-Stats-2023,@@#$month))::@3$3..@>$3=(remote(blog-ESS-Stats-2023,@@#$confromgrid)/remote(blog-ESS-Stats-2023,@@#$consumption))*100;%.1f::@3$4..@>$4=(remote(blog-ESS-Stats-2023,@@#$confromsolar)/remote(blog-ESS-Stats-2023,@@#$consumption))*100;%.1f::@3$5..@>$5=(remote(blog-ESS-Stats-2023,@@#$confrombatt)/remote(blog-ESS-Stats-2023,@@#$consumption))*100;%.1f::@3$6..@>$6=$4+$5;%.1f::@3$7..@>$7=(remote(blog-ESS-Stats-2023,@@#$solar)/$numdays;%.1f::@3$8..@>$8=(remote(blog-ESS-Stats-2023,@@#$consumption)/$numdays;%.1f::@3$9..@>$9=100*(remote(blog-ESS-Stats-2023,@@#$solar)/remote(blog-ESS-Stats-2023,@@#$consumption);%.0f::@3$10..@>$10=100*(remote(blog-ESS-Stats-2023,@@#$battnotfulldays)/$numdays;%.0f::@4$11..@>$11='(date-days-in-month (string-to-number (substring $2 0 4)) (string-to-number (substring $2 5 7))) Here's what I derived from the first table: - "Grid→Cons %": Percentage of consumption from the grid - "Solar→Cons %": Percentage of consumption from the PV - "Batt→Cons %": Percentage of consumption from the battery - "Sol/Batt→Cons %": Percentage of consumption from the PV or battery (in sum) - This is a good indicator of my level of independence from the grid. - "Sol/day": average production per day - So you can compare between months. - "Con/day": average consumption per day - So you can compare between months. - "Solar/Cons %": 100% would mean that I produced as much as I consumed. More than that resulted mostly in providing power to the grid. Less than that means that I had to get energy from the grid. On average, of course. - "%days batt<100%": percentage of days where the battery was not full at least once - This is a good indicator on how well I was able to charge and use the battery. In reality, this 13.01 kWp setup is producing a maximum of 8.8 kWp (figures from 2023-07-08, highest production peak). **** Home Automation :PROPERTIES: :END: I did not invest effort in another approach to get the PV into my Homeassistant setup. The conflict between Stiebel Eltron ISG+ and Victron Cerbo GX still needs to be fixed somehow. So far, the Victron VRM worked out great to monitor the system. However, there have been some changes that reminded me that I need to get rid of this cloud dependency somehow which allows me to cut lines to the Victron back-end in the long run. **** Minimum SOC :PROPERTIES: :END: Although I had other plans, I set still the minimum state of charge (SoC) manually. In late October, I went up from 20 to 40 percent. However, the generated energy dropped so fast, that I changed it in early November to 70 percent. And this is the value that made much sense over the autumn and winter so far. No single day produced that much energy that it exceeded 95 percent battery charge. Therefore, I might as well switch to 70 percent from mid October until at least February. And then I'll go down to my usual 20 percent in March which I keep as a power-loss reserve during the sunny months. No need for complex automation here. **** Dirt :PROPERTIES: :END: Although I've bought an extension to our garden hose so that I may clean up the panels whenever they get dirty, the level of dust and dirt is minimal. I did not clean them so far. Let's re-evaluate this at the end of winter and probably spring (pollen!). **** Going Dark :PROPERTIES: :END: We've had twelve days of total darkness on our panels in the first half of December. Snow covered the panels which reduced the amount of energy produced to zero. #+CAPTION: System overview from December 2023 #+ATTR_HTML: :alt Screenshot from Victron VRM with bar charts of various values. #+ATTR_HTML: :align center :width 630 :linked-image-width 1000 [[tsfile:2024-01-07T17.57.49 System overview of 2023-12 -- screenshots publicvoit statistics bwg.png][2024-01-07T17.57.49 System overview of 2023-12 -- screenshots publicvoit statistics bwg.png]] One of our neighbours even climed on his roof in order to remove the snow from the panels. I don't think that this is worth the effort and danger. I may have lost maybe one hundred kWh in total. There was one remarkable incident in December: on December 23rd, we produced almost 14 kWh on a single day. The next best figures in December were below nine kWh per day. It was a sunny day, yes but otherwise I don't know what caused this peak. **** Dark Months and Financial Situation :PROPERTIES: :END: It's a bit depressing when the power generation drops that much in late autumn, keeping us depending on the grid during the dark months. However, there is no way around until we do have some sort of decentralized way of storing PV energy for the winter. I don't think that this gets financially attractive for house owners any time soon. Until then, the grid is the storage. As long as I deliver more energy in summer than I retrieve in the dark months, that is fine. Especially when the money made by providing power to the grid is higher than the cost in winter. For 2023, it was a big financial loss: Because of high service fees and dynamic power prices of "Spotty Energy" I almost got nothing for my energy until 2023-08 because of choosing the wrong service. For about 3600 kWh (2023-03 → 2023-08) I got paid 65.64€ by Spotty. Then I switched my service provider to ÖMAG where I deliver energy. They paid 128.12€ (2023-08 → 2023-12) for just 910 kWh. That's a price of 1.8 Cent with Spotty and 14 Cent with ÖMAG. This Spotty adventure has cost us about 440€ loss. Let's hope that this improves in 2024. I do have some rough plans. If you live in Austria and you want to get power from our PV, [[https://7energy.at/][this]] could be a good option to follow. Only in mid-November, I got money from the Austrian sponsorship ÖMAG. It was late and roughly 3,000 Euro less than anticipated. I still need to find out what happened here. The lesson learned is that if you can't finance PV on your own without the support of sponsorship, you still need to bridge the expenses for almost a year. This is not good news to people who can't afford this financial gap. All in all, it's still one of my best investments so far and I'm looking forward when I - again - get more and more power from it over the next months.