Jump to content

New Micro grid system on BYD batteries going in @ 11.6kWp


Recommended Posts

This is a new system I designed around a Micro grid. Its a DIY job and its based off a Fronius Primo 8kW string inverter on the AC output of the Victron Quattro 8kW Inverter charger with a black start Victron 150/35 MPPT controller and 8kW generator for support. The BYD is a modular system, each brick is 4kWh and the system will initially be 16kWhs and in a few months will grow to 32kWh.

I will do periodic posts as I make progress in between the day job on this :)

spvd4-4.png

BYD bbox prem lvs.jpg

Link to comment
Share on other sites

I absolutely detest battling tech products and its even worse when they are not reliable. After 40 years in electronics, communications and IT I do not need or want a project that I have to rely on like Solar, in almost off grid and soon hopefully completely off grid, to be unreliable or to need to tinker on it all the time. This one needs to be completely hands off except monitoring and maintenance and tweaking to optimise. 

I love tinkering with electronics, old cars, computers, milling welding etc etc but won't want to with the solar system.

electronicsworkshop.thumb.jpg.e380bc87728ea62bf3872af84fcd8899.jpg

So my choices were to buy once and cry once. Been saving up for a few years to build this with plenty time to think over the design and how I want to build it and why. As to cost, just the major components like the Quattro, Fronius, 2 x BYD 16kWh packs, MPPT charge controller and 30 (24 Canadian 425W + 6 JA Solar 385W) were north of R300k

I will write up and post later on why on the component combinations and more about the Micro grid choice. 

Link to comment
Share on other sites

First some background, then I will dig into why and how: After having being an independent consultant to Eskom (Escam) for 10 years till end 2014, I do have some insights into that place. I will leave that there. Since the recent announcement by DeRuyter that the Coal fired Plant will get less and less maintenance (also look at the LHS and RHS graph from Escam data below) its clear we have collectively, a very big issue on our hands.

IMG_1443.thumb.jpg.da9297a09af6e10dc41a092b934acdf4.jpg

 

 

Then on top of all, there are those with connections that aim to have the Utility driven into the ground for their own financial gain. You may wonder how that is possible you ask? Well now if the dilapidated Coal fired Plant, and the new Medupi and Kusile Plant are not reliable or efficient, Escam must get power elsewhere, more places to bribe and corrupt. Look for this no further than the Powerships tender or any of the investors behind the new Green energy Plants. Or the latest round of coal procurement contracts at more than double past contract prices, makes you wonder.....

This brings me to my Solar PV system and the reasons behind that. I refuse to subsidise the Idiocy at Nersa and Escam. Talk about “gaan haal die bobejaan after die berg ne” I am stopping now to further fund the, in my opinion, blatant bribery and corruption. On to my system.  But first some more background.

 

Look at the access to the grid map here:

968578356_accesselectricity.thumb.png.e3a71c8c336f37e92e4666c65aaf22df.png

Now think about all the illegal connections to the grid and non payments (plain corruption and criminal theft) and all the new connections being made daily, both illegal and legal and its impact on your pocket, as we are funding all those illegal connections. I am gatvol for this. Not to even say anything about our own power (utility) independence. If your are dependant for heat in wintertime on Government, imagine how fooked you will be when they cannot deliver power, or worse, deliberately kill your smart meter’s ability to lo load new units….. Just think back a few years when during the holiday period, we could not buy prepaid units at all for a few days, it ended up in Parliment, for what that was worth. Lots of people will do anything to prevent that and do whatever asked, or just plain riot, burn and pillage. Look at our recent experience and that was not even a big event, only the ExPress going to jail.

 

See South Africa on here, our grid breakdown(not currently improving):

Our grid is 34% as Connected but working about half the time and 58% as works most of the time :D

global_20191212_fotw_fig1.thumb.png.a8bcbb02bfde6e60bc961b98bfb23fac.png

 

More to follow

….

Link to comment
Share on other sites

Posted (edited)

And as if on que, while I am typing this all up, a new round of loadshedding announced......... 🥶

Edit added: Eskom just explained to my earlier point that a unit at Tutuka (old plant) and Medupi (NOT old Plant) went down. And on top of that, the previous 2 units at Tutuka and  Medupi that were supposed to be returned today, are still down.

Edited by Sarel
Link to comment
Share on other sites

Thanx @87 Dream Will keep on posting. On the note of the BYDs, what attracted me was the reports on their reliability and very slow degradation. When I received them even the packaging was superb. I like the fact that its modular and there are no internal cables to be messed with at all, its all integrated in the modules. A lot of thought went into the detail and I appreciate things like that. Will post up some pictures of the BYDs later.

Link to comment
Share on other sites

Some more food for thought, since the new CEO started in beginning 2020, the unplanned outages at eScam has increased by ~4% :O, not decreased at all as lots of preventative maintenance may indicate. It is now 18 months since we were told that all the extra maintenance will be done. We had all the load shedding blackouts that were ascribed to maintenance. Not successful I would say…. But draw your own conclusions from the data. Also please validate and do your own investigations to prove or disprove the data presented, for yourself. You are welcome :) Also, these are just my thought processes, nothing wrong with any of the systems on here, just your compromises. My compromise was just on the money side as I chose to pay more to get exactly what I wanted. This is just my way of going about building my system around my needs and my compromises and maybe this can help people to make easier choices and understand why they make em. Lots of professionals on here with better or other ideas and all the more power to them :D Over the years I have learned a lot on here, consider this my contribution to the community, for what its worth. 

But now we talk Micro grid, aka mini me grid ;) or your own utility.

 meme.jpeg.909a6994044c04701d307edf28775e38.jpeg

I want to say again, I love projects, I really do and I work on lots of them at the same time. This Solar PV however cannot be a project, not so much as its the foundation for everything else, hence my choices here.

To start let’s look at the cheapest way to self generation and consumption, ie Grid tie systems. This is one of the fundamental systems together with Off grid and Hybrid systems. What are the problems of a Grid tie system? When the grid goes down, blackouts happen and then your Solar PV system goes down, by design. Just ask the People in Texas. The anti-islanding shuts down the Inverter, as it should from a safety standpoint. So without the grid it’s useless. Off grid systems add Storage (batteries) and are not connected to the utility grid at all. A hybrid system is a bit of both, ie Grid tied with Storage. Same problem exists for Grid tied Hybrid systems in that the Solar PV will disappear for some or most (depending on Inverter Architecture) if there is no grid and the Inverter cannot be connected to the downed grid. The system will dump all non essential loads when the grid disappears too.  Also your hard limit is the Hybrid Inverter capacity.

Enter the Micro grid. Essentially this means that your String Grid tied inverter must have a stable grid input of the required 50Hz frequency to function. The Inverter Charger and battery bank can supply this if the Grid tied string inverter is coupled on the output of the Inverter charger, Critical load position below or PV on AC output. If you connect it to the AC input, it will die with the grid. This is called AC coupling and can be used to inefficiently charge the batteries in the event of no grid. The inefficiency is due to the DC panels being inverted to AC and then again to DC with all their losses. There also may be battery assist where the Inverter/Charger will support the String Inverter by adding more power from the batteries to support the loads. There are all sorts of rules and caveats here and we can unpack those later.

DC coupling (Black start capability) is when the panels are connected to a MPPT controller direct to the DC battery side. Its way more efficient to charge batteries that way as there are no dc to ac inversion and back to dc. Effectively a Micro grid (there are larger versions of this called mini or metro grids that supply multiple consumers or homes) is a mini me Eskom, independent of Eskom. You can under certain conditions in a Micro grid, expect to get extra kW support from the batteries in addition to the String inverter’s power output, up to the limit of the Inverter thus adding between 1 and 100% of the power from batteries, but with lots of ifs and buts and caveats in there.

microgrid1.thumb.png.69e01a8baa232afefa8d3ce4639886c0.png

PS: I am not affiliated with any company at all, its merely convenient to use my own system as example. I paid full retail for.

My system is a Micro Grid with the Victron Quattro 8kVa and Fronius Primo 8.2 single phase string inverter on the Quattro AC output, thus creating my own Micro grid. I follow the 1:1 rule for adding a Fronius on the Quattro’s output. That means you are not allowed a String Grid tied inverter of larger capacity than the Quattro, in my case at 8kVa, hence the Fronius 8.2 Primo. I am still going to be connected to the Utility grid in the meantime. I will go Off grid when my BYD storage has been expanded with enough capacity to last a full dark Winters day and night.

I am using both AC (Victron & Fronius) coupling as well as DC coupling (MPPT charge controller on DC Bus) to allow the sun to charge and start the system in case of no grid and depleted battery. And yes you are correct, I can wait for the grid to return to start the system or run the generator. As explained, I do not want to be dependant but rather independent from the Grid and also do not want to mess with the system. Black start capability will automatically recover the system as the Fronius will need a grid at 50Hz to start, and take up the load. The DC MPPT system can charge the batteries even with Grid down and batteries depleted allowing the Quattro to boot and provide the grid again. 

More to follow....

Link to comment
Share on other sites

18 hours ago, Sarel said:

And as if on que, while I am typing this all up, a new round of loadshedding announced......... 🥶

Edit added: Eskom just explained to my earlier point that a unit at Tutuka (old plant) and Medupi (NOT old Plant) went down. And on top of that, the previous 2 units at Tutuka and  Medupi that were supposed to be returned today, are still down.

Come to think of it now, I should have titled this thread "Bucle up Buttercup" 🥸

Link to comment
Share on other sites

Basic 2P BYE 16kWh system standing. Monday we should start with the panels. Its an old IBR roof so no modern brackets will fit 😏

These are the internal connector blocks for when you stack the modular 4kWh units together, no wires yay!

wirlesss.thumb.png.c0fff67e8f8e90cf5a91bc3cddd89b63.png2x2pack.thumb.jpg.c197cea195c54dc65de059bf5f056ded.jpg

Link to comment
Share on other sites

Sarel these BYD batteries are a thing of beauty...I will only add that the Tesla Powerwall still wins on space. However, the Powerwall is not this size in terms of kWh. 

Your discussion on the mini grid is really interesting because I honestly think this is the future of home Electricity. The big Grid in my opinion will disappear all together. 

Please keep the updates coming it's a cracking project ☺️

87

Link to comment
Share on other sites

Thank you @87 Dream Will do. A few advantages for me of the BYD over the Tesla:

  1. You can control the BYD, You cannot control the Tesla, only Tesla can do that
  2. BYD is modular from 4kWh to 24kWh per tower as a single tower (6 modular units of 4kWh)
  3. You can parallel up to 64 towers with 4 units & 16kWh ea for a total of 256kWh ( I do want to see that DC busbar system 🤑)
  4. Apart from the cables to the DC bus, there are only the comms interconnects between towers
  5. No DC cables to manage or control modules, only one BMU module per system
  6. 250A discharge per 4 unit tower with 360A peak
  7. No cabinet, or cables to build the stack and its IP55 rated
  8. Cobalt free LithiumIronPhosphate with 10 year warranty
  9. Good value for what you get, in my opinion
  10. Adding a new module takes seconds as there are nothing to do but plonk it on top of the existing.
  11. Complete assembly takes less than 5 minutes

See we cranked the volume to 11 there 😬

Link to comment
Share on other sites

Posted (edited)
2 hours ago, 87 Dream said:

Sarel these BYD batteries are a thing of beauty...I will only add that the Tesla Powerwall still wins on space. However, the Powerwall is not this size in terms of kWh. 

Your discussion on the mini grid is really interesting because I honestly think this is the future of home Electricity. The big Grid in my opinion will disappear all together. 

Please keep the updates coming it's a cracking project ☺️

87

One thing on the Micro grid, same old SSEG rules still apply. No wheeling allowed, all energy generated must be consumed on the same property so no selling to next door etc etc. But its got, in my opinion, lots of advantages. Money not being one of those sadly .... But energy independence is tops! 

Edit to add: There are lots of global initiatives to combine the grids. SSA region (Sub Sahara) already mostly interconnected grid. Alas, Buggerments everywhere will legislate it in existence always.

Edited by Sarel
Link to comment
Share on other sites

Have to agree with all eleven points: even add an extra one: makes better financial cents so to speak. The Powerwall is a crazy lot of money to get into SA. Last time I looked was R160k odd 🥴😪

Yes great point about the dangerous Chemistry used in the Tesla. If I had my own way I would love to get a LTO (Lithium Titinate Oxide) build going but the cells are just crazy money at the moment. But at 25k odd cycles & a very very stable chemistry this is the ultimate battery no matter what. Also can do 4C charge & discharge rates. 

On another point yes the Efrika Governments will try their best to legislate against us. The only way is to get grid independent... Therefore they can't say much.

87

Link to comment
Share on other sites

Important difference between the Powerwall and other systems are that Tesla made it an AC battery with build in charger. (120 or 240V) Its a 5kW system continuous. It uses NMC chemistry and is liquid cooled, so derating for temps are less than others. Its a different beast to integrate into our solar systems vs the normal DC coupling systems. Nothing wrong with that, just a different way to the end. And it Expensive, you can buy same capacity for less from other brands locally.

Tesla’s virtual grid is just awesome tho. And hopefully soon Elon will release vehicle to grid technology as well.

Link to comment
Share on other sites

I think in tesla products aesthetics are given higher priority than other companies. In low crime countries the powerwall is marketed to be installed outside the garage where everyone can see it. For the cost of one 13kw powerwall you can get 30kw pylontech/dyness/freedomwon batteries. The marketing for it is very good but I had to come to terms with that I'm just not the target market.

Link to comment
Share on other sites

For those who wants to know, the tech specs for the BYD LVS batterybox. (Low Voltage Small)

My config is 16kWh split over two towers in parallel. This will be expanded by October with 16kWh across the two towers. Later a third tower of 16kWh will be added for 48kWh storage.

Each of the 16kWh systems can support a 12kW load, not that I will be getting close to that with the 8kVA Quattro. Supporting a continuous load at 25c will consume about 6800W at 48V, so 142 Amps.

Split over two towers in parallel, this will result in 71A draw per tower max. The Quattro can support a 1 sec peak of 16kVA and draw 284 Amps total. Still comfortably below one towers peak load capability at 5 sec and 360A.

The initial 16kWh total (8kWh x 2) can support 260A loads, so also safe.

7C857011-011D-465D-A4ED-2F4B8F5A7532.thumb.png.3cfeed2e57441544019e29a1b8465f77.png3CE435AF-403F-4D99-9C87-CF6DF21EED77.thumb.png.0d80c732f8a966069fe1dda9675b2a67.png4FA777C0-1928-48F3-BA74-65C4A9B78944.thumb.png.d791294202c2b63f18a6814a31e56c5e.png

Link to comment
Share on other sites

Posted (edited)

My setup is in a very old house. I still have the original copper Ryalco geyser. This thing is fitted with a ~4kW element. I am not planning on replacing this for various reasons. The current setup in preparation for the Solar install is as follows:

A timer and a Sonoff switch combined that controls a SCR controller driven by a little adjuster that changes the duty cycle of the SCR from 0 - 100%. I will leave out the gory technical details for now. This allows me to control absolute times with the timer that the geyser must be switched off. The Timer has a NO and NC contact set. I am using reverse logic (to help ensure that if timer fails, we have a chance that the Normally Closed contact will remain closed under no power condition) and the Sonoff can still control the geyser heating. The timer and Sonoff only controls the mains feed to the power supply for the control gear, so a few milliamp's at best. The life and reliability of the Sonoff and timer is improved. When the timer is active, ie. closing the contacts, the SCR control voltage is interrupted via the NC contacts, now open (24V for a 4-20ma control.) Thusly ,when the timer switches on, it inhibits the power supply to the PWM and no heating. So I am setting the timer for when the geyser cannot be warmed. 

The SCR is a zero point crossing time proportional control ,so much gentler on the element. It will duty cycle switch power to the element.

438561970_ZeroCross.png.6f4f9f55fb0a6dcf920877033082e7c5.png

scrcontrl1.thumb.png.cc40416c0a892c124863b7765f540ad4.png

SCR proof of concept video

Now this allows me to control and dial in the energy consumed from 0 to the full kWh of the element. Each cycle of the mains will still consume the full kWh but on average the load will be what we set on the dial. My timer enabled the geyser at 13:00 and switched it off at 16:00. My base load was ~600 Watts and the 4kW element was fed about 2kW via the SCR control. There were most likely a voltage sag at that time or the element is not really a full 4kW unit, likely consuming only 3.5kW or so. This control is done to allow me to tweak my Fronius and Quattro load and allow better control over charging and keeping my feed in to 0.

Herewith my consumption. (the peaks are from other loads.)1534014308_Screenshot2021-07-26at08_14_31.thumb.png.7909673a2c2c1737b479f2f646601c5b.png

I will describe in a future post how I will use an Arduino to control the heating periods and then the duty cycle based on a few parameters form the Solar system.

Edited by Sarel
Link to comment
Share on other sites

Some more background on why this has to happen in this way.

Since I am planning my setup to eventually migrate off grid, all loads are critical loads. That is why I selected the products I did. The one issue I have is a 3Ph compressor, surface grinder and tool grinder that needs to be fed. I am on a single phase supply. I generate my own 3 phase but it is a lossy way to do that, good enough for now. Most other tools like lathe, mill ,welders etc are single phase. Their draw varies between 1.1kW ea to 4.5kW for the welder. The 3PH compressor draws almost 6kW running, startup will be way more. I may have to stick that onto a VFD to manage its inrush as that will kill my Inverters, we’ll shut them down. Most of my workshop work is done during daylight, seldom after sundown. After dark its mostly lights and hand tools or small power tools used and infrequently air tools and the air tool work can be moved to daylight hours.

All the house items are mostly now around 2kW each, even my geyser thanks to the PWM control via SCR. My stoves are all Induction tops and I use a whistling kettle on the induction for boiling water. Air fryer, microwave, dishwasher, espresso machines, ovens etc etc are smaller loads and easily managed in serial. The Cerbo GX with display will be installed close to my bedroom, all mobile devices will have the ability to monitor the loading of the system, also via Home Assistant. I am installing a permanent wall mounted tablet in the kitchen to monitor the loads while preparing meals etc.

Most heavy loads (1.5kW or more or consuming more than 1kWh) have already been time warped (shifted) to daytime to get the family into that routine. At some later point I will have to parallel my Quattro for 16kVA or will have to go 3 for 3PH supply and will have to get a second String inverter as well with more panels to support. I my have to change the Fronius string inverter to a 3PH unit.

 

 

Link to comment
Share on other sites

Posted (edited)

Methodology of planning my system.

PS: I will add still more on the Micro grid later.

I decided to start with the source, the Sun. You may search for your town data on the web.

390623894_pretoriahours.thumb.jpg.33b1123b65be04f849da78151f3422bf.jpg

The above is deceiving as the months start on July instead of January :D The grey area, for Average daylight and the row  below that for cloudy hours.

 

Due to my horizon, houses  & trees to the east, trees to the northwest and west, I will lose about 1-2 hours of sunshine during winter months and some during summer. So I limit the available time from 08:30 - 16:30 per day, ie 8 hours sunshine per day.

 

The below is a map of the potential to generate PV power. We getting plenty.

1583564934_SPVG_opt_ZA.thumb.jpg.2a19e73cd16ab5f1f9ee8a872ddb9532.jpg

If we look at the global map for sunshine and compare it with the cloud cover map below that, it is clear with good correlation, cloud cover is a big factor in limiting the solar potential. 

World_re.jpg.f62a903e70780abcebc7760860774963.jpg

Annual-Aqua-PM-globe.thumb.jpg.d9ec27fb4beb597ec859c9fd676503d4.jpg

Now we look at sunrise and sunset together with the Azimuth and Elevation of the Sun for 26 June. These lines are where the shadows will fall between sunrise and sunset. At the time of 08:00 the sun will be 12.22 degrees above the horizon. Still behind some trees etc for my panels. You can grab this info for your location and model the same. https://gml.noaa.gov/grad/solcalc/630800935_Screenshot2021-07-15at20_44_35.thumb.jpg.5f563065057459f825e76e07730d278b.jpg

Now that we established more or less how much sunshine we can expect and where the sun will be relative to our panels, we can move on to the next step, energy consumption.

I had an energy meter installed a long time ago.

1031132625_DownloadUsage.thumb.png.50b16f90c14cc9128b940f84c099cfb0.png

918783863_DownloadUsage.thumb.png.3f7df2bb78d73a453ebece4b99ced712.png For 2020 the consumption is much higher as we started working from home. We can see the demand in kW as well. Between the demand and consumption we can see the patterns of our consumption and can base the capacity planning on real data. In my case I need at least 6.5kW and sometimes this can almost double to more than 12kW although that is for short durations only, 30-40 minutes mostly.

 

Now we have established I need to support up to about 12kW load but my average daily peak is mostly between 500W and 7kW the 8kVA Quattro should be sufficient. To be able to support higher loads than what the Quattro can deliver, they will be time shifted to daylight hours only since the String Inverter will help support the other loads like Geyser etc. It can supply another 8kVA on top of the Quattro.  I will explain my Geyser method later in more detail. When looking at my daily consumption, worst case from 18:00 to 06:00  is consuming about 42kWh. Some of this load will be shifted to daytime so leaving me with about 30kWh during the night. This is worst case during winter. Summer is much much less. So now we know how much storage I need. I settled on initially 16kWh and will later double that to 32kWh. When budget allow we will add another set for a total of 48kWh in 3 parallel packs.

 

Next we will look at the required Panels and their capacity.

Edit: removed duplicate image

 

Edited by Sarel
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...