 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : How to employ smaller storage systems for best efficiency.
Page 1 of 2   |
|||||
| Author | Message | ||||
wiseguy Guru  Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
I will apologise in advance for the following over verbose text and mix of simplicity and probably overly technical information. There will be a percentage of readers who will get what I am doing as it may also solve their particular similar issue. If anyone has other possible or simpler solutions, I would be happy to hear of them. I currently have a 3.6kWH battery pack (actually a super capacitor storage) which is way too small to go off grid with. I have also agonized over how to use my home brew inverter to power the house when the roof solar stops feeding into the grid in the evenings. I intend to increase my storage to 5 or 10kWh once I have this system running. I am inner city, I have 13.2kW of solar generation (44 x 300Wpanels) & a 10kW 3PH Fronius grid tie inverter. I have not received a bill for electricity since install in 2019, but my feed in rate was recently reduced to 9C for the first 40kWH and then 5c after that. We feed back ~80kWh/day for ~ 6 months of the year, so for the first half day of a metering period we get 9CkWh then the balance of remaining days is at a paltry 5c. Our feed in rate in 2019 began at 18c/kWh, but I expect I will have to pay towards my power this year. So a scheme with any efficiency gain will help to minimise my bills. Edit: I forgot to include some important information in my first post, most of our night-time consumption occurs on 1 phase, so I really need to reduce that consumption to zero. In time the other 2 phases consumption will be added to the first phase so with net metering (which we are on) all draw from the street should be able to be negated. Of course if I need some power from the street it is inevitably at the peak tariff and it is donated to me for ~ 40c/kWh. So in 1 hour for 1kW of feed from the street I need to put back 8 hours of 1kW at 5c/kWH to be cost neutral. What a lark if you can get away with it. This smells of a sharp practice where else can you receive what seems like an ~800% markup - on what is really an essential service legally? So, plan 1. A first draft messy scheme goes like this, turn off the street power, disconnect the dedicated solar inverter. Insert the plug that shorts L1,L2 & L3 together connected to the inverter output. Start the inverter and throw the switch of the 3 phase plug to single phase inverter feed, yay we have house power on all 3Phases until the batteries go flat. Of course before the solar starts to feed again the next morning or if the batteries have gone flat during the night, we want to have reversed the previous steps. I am sure you will agree this plan has about zero merit and really sucks (technical term). Plan 2. Recently a colleague has been playing with a different plan to achieve the above function with simplicity as the key. His concept as my colleague described to me is that he reads ac voltage and current from the street to determine magnitude and direction for his control system. Positive power in the text below refers to power drawn from the grid, negative power is injecting to the grid. If street power is positive to the house,the intention is to feed power into the house power from your battery storage using a second grid tie inverter to null the positive draw from the street up to say 4kW. So at night the system will power the house requirements. If street power should become negative (= roof solar producing more than house requires) now recharge the battery storage system as a priority before returning any excess energy to street. This scheme is AC coupled and completely automated needing no active (manual) switching at all. Below is a more technical description for anyone interested in the more technical aspects. From the power unit in the switchboard, he transmits packets of data for V, I, W & VA from a local ESP32 to another ESP32 away at the receiving end. The scheme also involves the second off the shelf grid feed inverter (with current approvals) which is fed from a 48V storage source through a 48V to 200/300/400V(as required by the inverter) DC to DC converter at up to 4kW. The HVDC output of the DCtoDC will have a current sensor to be able to control the output power to the inverter (it pretends it is a solar string). The output of the inverter is connected to house power (if under 3.6KW via a convenient 15A 3 pin outlet) The Remote ESP32 creates a control voltage via PWM which is usually zero when there is no positive power required and the 48/300 converter does nothing except regulate the 300V or sleep. If power is now positive it ramps up the PWM control voltage of the DC/DC towards maximum (3,4,5V?) until we are feeding either our maximum capability of power or until at some point the grid power is now zero. Of course with a 4kW inverter we can in theory null a power draw of up to 4kW at which point we stay at our maximum and the normal grid just supplies any remaining deficiency. So Max PWM is max DC/DC & Max Power. Bells & whistles: if PWM is zero for x minutes it means please sleep the DC/DC. I envisage our control voltage from PWM to be say 0.5V - 3.5V = 1W to 3,4,5kW - what ever DC/DC or house feed inverter is the limiting bit. Therefore below 0.5V = negative power already so DC/DC = sleep. I believe an ac coupling of the pwm output and a watchdog that says pulsing (pwm) is ok please run DC/DC, but jammed at zero or maximum with no PWM = lets not do anything (sleep the inverter) for a failsafe approach. The isolated DC/DC approach is required as most transformerless inverters are essentially live to grid at the solar terminals so for safety at the 48V side better to be well isolated from the mains. I am currently working on the power measurement device at the street mains to determine direction and magnitude of street power activity. We dont need to know the true power value accurately as long as it has linearity and headroom the servo loop just nulls whatever Power it sees. We dont need a precision calibration of the unit = if needed we can calibrate the numbers returned with an off board correction constant in our software. One issue which is a bit more difficult is phase shifting I relative to V to be very close to being in-phase for a resistive load, we can do it in software too but maybe better to do that one on the measuring board as a digital calibration. I am also working on a 48V/300V DC-DC at 3kW minimum, current controlled converter which might sound a bit scary but I like living on the wild side and approach these projects with “failure is not an option”. It must also be within the capability of a home constructor, so I have my work cut out for me but I think I have a novel solution that will be tested in a week or two, currently on holidays remote from workshop…... My colleague employed a deadband when controlling power generation that says if I am within a band of a few % just (think a few watts) of null be happy until power increase or decreases beyond the few % band before adjusting again. When he reads voltage at the street feed is below ~ 180VAC he stops generating also. The finished box with our bits may not have approvals but legally it can plug into a 3 pin plug and the current sensor clamp clips onto the street feed lead, so I don’t envisage any legality problem ? Note this scheme is purely aimed at minimising our power consumption from the grid - we would probably be accused of an activity that is considered by the supplier as not acting in the right spirit of things. However the energy companies are stealing our power for a few paltry cents for our energy (if youre lucky) and charging us 40c+ per kwh at peak times when we need it - daylight robbery. They arent in the right spirit of things either, a new solar farm opened in a township out of Adelaide recently with much fanfare - I bet they get more than a few cents per kwH ? During the day if ever we need more power in the house than the roof solar creates, the second inverter can inject up to 4kW of energy transparently as required. During a few days of next to zero solar, & 1 hour before off peak cheaper power finishes we can charge or top up our storage bank using cheap energy and at least minimize (the first 3.6kWH) in my case of higher cost energy. When we have oodles of solar the first priority before sending any excess solar to the grid is to charge the battery up, we can now use the power and direction unit to null the negative power now to the street until the batteries are fully charged up. If the batteries ever go flat, if the second house feed inverter fails normal street power resumes/continues without any fuss or interruption. The last part which is a great bonus is that any reactive power is supplied from the grid and not from our battery storage so we don’t have to pay for the reactive components of our house power requirements ! I intend to build a small trolley for the storage and extra Fronius inverter (I bought a working 4kW Fronius Primo for ~$400 total outlay) and have currently installed a 3-pin plug to the Fronius for my experiments. The trolley will also house my home brew inverter which will be employed during a total grid failure (blackout via a couple of switches). I will share results and schematics software etc after it is proven to run successfully. Edited 2022-12-26 21:09 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
renewableMark Guru Joined: 09/12/2017 Location: AustraliaPosts: 1678 |
I know I am not answering your questions here or even helping at all, but this is exactly why I thought FK the grid, I'll just remove myself from it, throw some dollars at storage and be done with it. Doing off grid you simply have to pony up the dollars for batteries. I think your energy provider would have a fit if they read this (and you know it). Cheers Caveman Mark Off grid eastern Melb |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
I hear you Mark but when the solar went in there was a decent feed in tariff and no hint that it would ever reduce to zero - which a lot of energy suppliers have now done. We produce more energy that we could ever use, so being able to dump heaps of it onto the grid and get a reasonable return was all part of the strategy (stragedy?) and cost justification to install it. It is quite annoying that now we will have to start to pay for power despite exporting shiploads onto the grid almost for free so those toads can make a killing reselling it for a big profit. I think I will probably just buy a heap of storage eventually and do as you suggest. The above I guess was a way to minimise my consumption to make the miserly feed in tariff still work for me at minimal cost and get free reactive power to boot lol payback! Edited 2022-12-26 23:01 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| nickskethisniks Guru Joined: 17/10/2017 Location: BelgiumPosts: 458 |
That's a nice project Mike! I'm always ready to tinkering about this strategy with you, I would do it the same way. I'm playing with this idea myself for a long time, I'm completely off grid but at my parents I want to do this kind of project. In EU you can of course buy such, home batteries (government funded systems), but what's the fun in that? They (parents) have also 3ph system, and one question I asked myself is how the consumption is measured? For example if you inject 3kw at 1 phase but use 1,5kW on the other phases, what then? Is it like the average 0kW so no costs? Or are you using 2*1,5kW (3kW) and injecting 3kW so you have to pay the difference? My idea was to use the existing inverter to be fed with an isolated DCDC (CV-CC) by night or clouds to compensate and regulate consumption at 0W. Just add 2 diodes, one in series with the solar panels and 1 in series with the converter for safety, it's 3-400V so the losses of the diodes are marginally. They do have a digital meter, so I can just extract all data out of that meter with a uC or rpi for example, that way I can do calculations and instruct the dcdc. I was thinking a lot about how to detect the direction of the current (using or injecting), so 1 current meter or actually 3 for 3ph. But if you double the current sensors it would be easier, no? And probably a 3ph gti will sent equal power thru all the phases, right? so only 4 sensors needed? I was more aiming for 2-400W, just for compensating idle power and move on from that. I will always be attracted to the multi kW modules. Last year I was doing a lot of research in high power converters at work besides the managing function I have so not the best combination. At home I was doing some experimentation in a TL494 based cc cv converter, specially the control loop but on a slow pace. An altered version of the poida mppt software might be an option as well, maybe not fast enough though. I know I still need to update the energy recuperation topic I started here but work kept me from making progress, the tl494 solution worked for a certain range but had stability issues with current control beyond that. I didn't knew all what I was doing at the start and I wanted fast result... There was also a ground issue in the pcb that was regulating everything, debugging took to long so I put it on hold because I lost motivation. I'm taking some work related actions for improving my health and free time and I hope I have more free time and I can share progress again in the future. I did already wind a prototype transformer for 5kW at 20khz, with 2 U100 cores, H bridge topology.  Edited 2022-12-27 08:08 by nickskethisniks |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Nick, With regard to 3 phase use, I am creating a 3 phase monitor for V/I/W & direction which can be used for single or polyphase measurements. It is based on the ADE7754 of which I have quite a few in my stock. I have plenty of very accurate current transformers, but an electrician would have to disconnect terminals to slip them on the wire. So instead I will use the lesser performance snap on types and figure out how to deal with any resulting phase accuracy issues. "For example if you inject 3kw at 1 phase but use 1,5kW on the other phases, what then? Is it like the average 0kW so no costs?" My understanding is that your above quoted line is correct, no cost. I believe in Australia most power is using what is called "net metering" where the total energy consumption of all 3 phases is compared to the generating return of all 3 phases and we pay (or are credited) for the difference. 3Phase inverters though return equal energy on each phase during generation. In my case I am often dumping ~9+kW for extended periods onto the grid. The problem as I see it, by utilising the existing grid feed inverter, is lets assume it is night, if I try to inject energy for 1kW consumption on only one phase, the other phases will feed to the grid power I dont want to lose. The system I am trying to create injects almost to the watt the actual consumption at any given point in time so real consumption from the grid is zero. However, my injection is intended to eventually be the exact sum of the 3 phases, fed back on the Phase with the worst consumption. The injection back to the grid will be the value of the sum on two phases with small consumption added to my injection that reduces the third phase to zero ie that phase is overcompensated and feeds back to the grid the sum value of the other 2 phases only. It may be possible to control the existing 3 phase inverter in some manner that achieves a similar outcome, but it hurts my head a lot to try to figure out what to do to achieve it. I can see a solution that I am able to create that only hurts my head a bit. The Fronius brand is held in very high regard with proven excellent long-term reliability. A bonus is that good 3-4-5 & 6kW inverters are being removed in bulk to install hybrid solar/battery/inverter solutions as owners try to minimise their reliance of the grid. FYI in South Australia we have had a few instances of total grid blackout in vast areas for up to a few days and we are not even in Ukraine! (Thank the Lord...) So cheap good brand inverters are becoming quite readily available, my $400 investment is still under new unit warranty and should live in my intended application for 10+ years. I too began with wanting to compensate for ~ 500W typical night loads but I am confident that I can achieve well in excess of 1kW and with dual or 3 stage in parallel maybe up to 4kW easily. That way when sun is behind clouds during the day and my consumption is more kW than the solar can provide it can be supplemented by the new system instead of drawing from the grid, hence wanting more than 500W. BTW Nick I have never played with U-Cores despite their high-power capability for three reasons. Are bobbins readily available or do you have to print/make your own ? The second is that I have large quantities of ETD & EE style cores up to EE70 and bobbins in my drawers already, so they usually end up in my designs. The third is availability and cost - where did you obtain yours ? Edited 2022-12-27 12:20 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
I just found an interesting clip, on new vehicle storage systems by Mercedes here I reckon I can see a defamation case coming on. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418 |
Wiseguy, I made a simple as can be drawing on google docs. here where we might be able to edit. I need to work out how to let anyone edit this... this is what it is NOW.  In Red, I want to show what you want it to be. can you 'splain to me what you propose? I will then modify drawing to show this. wronger than a phone book full of wrong phone numbers |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Peter, Lets talk single phase for simplicity An energy monitor added which reads street volts and compares it to current from a current clamp, installed on street supply wire between the street and the meter box. This will measure energy and direction ie drawing or feeding back energy. The energy monitor has a wifi link to a remote device. There is a 48V battery storage block to add and another block which is a DC to DC converter (48 - 300). Another block is the second grid feed inverter that is directly in parallel with the main inverter and its panel input is connected to the dc/dc converter. The remote wifi receiver controls the DC/DC from 0 to Max via a dc control voltage pwm. Ok having written that I am now getting a sense of a gentle hint that maybe I should have drawn an accompanying image lol wait a little while and I will try. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Here is a picture that I hope is readable and represents the proposed system add on in red.  When power is attempted to be drawn from the street an equal amount of power is injected from the second grid tie. So in effect it supplies the required power instead of the street ie street power does not feed the house, our own solar does during the day. Our second grid feeds the energy when there is not enough solar Apologies everyone a picture is worth lots more than a verbose description I can see the error of my ways..... Peter I am happy if you want to add the changes to your above creation so others may edit further as required. Edited 2022-12-27 21:46 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024 |
Not to worry, I was reading that and understood most of it.  Option 1. I used to have 2 phases in from the grid, the previous owner had an electrician wire up a generator input with a change over switch ON/OFF/ON and it would join the 2 phases in the switch and supply the entire house and shed with single phase. I did see a small box called "Catch Power Gen2" to capture power when a GTI is exporting to the grid and use it for the HWS......I'm sure your method will be much higher level. Option 3. or what ever number it ends up.......Do As Mark said.  Cheers Aaron Off The Grid |
||||
| poida Guru Joined: 02/02/2017 Location: AustraliaPosts: 1418 |
WG: thanks for the drawing. Now I can see what the idea is. wronger than a phone book full of wrong phone numbers |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Nick, with regard to your idea of diodes I have probably 50 x 1200V @ 50A stud diodes in my stuff. So diode or-ing of power is certainly feasible - but for a braver man than me. Playing with the 800V (I think that is my string voltage open circuit) @ ~20A scares the bejesus out of me and I intend not to go anywhere near it - EVER! I will play with kilowatts at up to 3-400VDC (lower DC voltages are preferred) but DC voltages that high are real scary when something goes wrong and the energy just doesn't stop from solar unless it finally becomes shorted or the sun goes down. Until the sun goes down there can be a 13.2kW plasma fireball the size of a softball tracking the wiring back towards the panels as it vaporises & sets fire to everything combustible nearby. This may be slightly exaggerated but is the image that comes to my mind when I think about it, the next thought is of Darwins principle along with, don't go there! After writing this I wondered if it was a bit over the top so I went to youtube and looked for solar arc, this is the first (and only) video I looked at, it has 4 panels (I have 44 on my roof) and demonstrates what a much lower energy solar plasma arc might look like link so think carefully about it before fiddling ! Edited 2022-12-28 08:53 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| pd-- Senior Member  Joined: 11/12/2020 Location: AustraliaPosts: 122 |
Your description sounds pretty accurate to me I put together sum photos for a solar panel fire course a while back if you are going to play with this stuff then a rapid shutdown system and running the pos & neg in separate conduits that are far enough apart that the arc will blow out are a good idea I had a look at the modbus docs for the fronius inverters, i think you would need to use three single phase inverters if you wished to be able to very output per phase  |
||||
| Godoh Guru Joined: 26/09/2020 Location: AustraliaPosts: 458 |
Wow PD, that is one scary photo. That is why I included circuit breakers as well as isolators under my panels that are on a ground mount. They are on a cross beam under the timber panel frames. I have a 30 metre underground run to the shed so wanted to make sure that if a fault occurred then the underground run, that becomes a run through the shed was protected. I am only running around 70 volts as well. Those very high voltage DC runs on grid connects scare me. I have seen panels that have burnt where they shorted between the tracks on the panel itself. Great big burnt patches. Pete |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Paul, Thanks for the picture confirmation, yep that's pretty much the image I had conjured up in my mind. Yes, the 3 Phase inverters I have looked at to date usually balance the power delivery equally for each phase. I intend to use the single-phase approach on the most heavily loaded phase which will keep me happy for minimal outlay. I very briefly considered 3 single phase inverters sounds like a long pay back period - you could go broke trying to save money...... If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
phil99 Guru Joined: 11/02/2018 Location: AustraliaPosts: 2135 |
Our local CFA brigade has attended one just like the photo. 10 x 300W panels. You don't need hundreds of volts either, a loose connector on a large 50V DC emergency lighting system did the same thing. With no half cycle zero crossings to interrupt the arc it just keeps going. One reason why welders prefer DC. |
||||
| pd-- Senior Member Joined: 11/12/2020 Location: AustraliaPosts: 122 |
With the 48V/300V DC-DC converter im thinking that it just needs to supply the 300v to the grid tie inverter as a constant voltage supply your bidirectional meter could then talk to your fronius inverter via its rs485 port and control its output. The fronius can be programed in fail safe mode so if it looses comms with your bidirectional meter then it will reduce its output to 0 Thats assuming that you are planing on using a fronius for your 4kw grid ti inverter Its just a bit of a twist on how things work hear with my AC coupled off grid system. |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Hi Paul, I am quite unfamiliar with what Fronius information is available really. Is the protocol easily obtainable somewhere ? I assumed it probably wouldn't be, the other reason for making the 300V a programmable current out was that the concept is then not tied to a particular brand of inverter. However, in my case I will be using a Fronius so it still could be my option - I really like their good name and reliability despite being annoyed at how hot they run - gotta be some inefficiency somewhere in there despite their reliability ! Edited 2022-12-28 23:03 by wiseguy If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| nickskethisniks Guru Joined: 17/10/2017 Location: BelgiumPosts: 458 |
Ok, 800VDC is not something for home use for me ;) At work we play with 3kV, but that's another domain. You are totally right, I was totally amazed when I was playing with 2 panels in series one day, just making arcs, only about 75VDC and not even in full sun.  That's one main reason for me, to not go over a 48V nominal battery system! I would also go for the extra gti. For the cores, I bought them about 4y ago for my choke in the ozz inverter, 10€ each with 10 halves at digikey. Prices went up since then, a lot! Most of my components I buy here: https://www.tme.eu/be/nl/details/u100_57_25-3c90/ferrietkernen/ferroxcube/ Bobbins are a problem, I made some out of ABS with a 3D printer, but you can't use them above a certain temperature, 70°C constantly should be possible. Making something out of FR4 material can be an option. I bought a few I100 pieces too, to make monster transformers but those etd and EE cores are probably better with EMI and directly mountable on a pcb as well. Edited 2022-12-29 06:15 by nickskethisniks |
||||
| wiseguy Guru Joined: 21/06/2018 Location: AustraliaPosts: 1156 |
Nick, I was eyeing those exact U cores from Digikey yesterday and downloaded a Ferroxcube data sheet for them. They are an eyewatering $49.22 ea for 1-9 pieces which does not sound like a huge bargain compared to a working 4kW Fronius for $300 + freight I bought a couple of weeks ago = to 6 U cores.... The only bonus is that digikey pay for freight if orders >$50 so if I order 2 or more freight is free. It has made the EE65's & EE70's I have to hand suddenly look a whole lot more attractive again. If at first you dont succeed, I suggest you avoid sky diving.... Cheers Mike |
||||
| Page 1 of 2 |
|||||
 Print this page |
