A few weeks ago, I embarked on yet another journey. I noticed a TMC thread about using an EV to supplement a micro grid; 4th link. That is, get power and energy out of an EV battery in order to lengthen the time that a micro grid (such as a PowerWall) might operate in the absence of sunlight and the absence of the grid. That can be done using the EV's 12v system. Those 12v systems are charged from the EV's traction batteries through a DC to DC converter which reduces the traction battery's voltage from 300-400vdc to just over 14vdc. Those DC-DCs can supply at least 1kw, more likely 2kw. Right now, I'm focused on getting about 500w out. I have been working trying to get my standard Enphase M215 microinverter to convert from12vdc to 240vac. I am using hundreds of M215s, one for each of my 60 cell, ~245 watt panels. Alas! Most grid tie microinverters expect to be powered from PV panels and are optimized for that source with something called MPPT. MPPT seems to prevent the inverter from working from battery power. The jury is out as to whether M215s can be used for that application. Meanwhile, I'm going to try a grid tie micoinverter that is intended to work from battery power. An example is the first link. The differences between the one used by the TMC experimenter and the one below:
1) The one below has 12vdc input; the tested one requires a voltage booster.
2) The one below has 240vac output; the other, 120vac.
I was led to believe that M215s might work from a non-PV power source by a college project to make exercise equipment supply grid power. The 3rd link.
In order for EV battery power to do you any good, you need a micro grid. Tesla PowerWall is a very good micro grid. However, Tesla discourages use by refusing to sell, refusing to install, and refusing to support. A seemingly good alternative is the 2nd link.
A single battery PowerWall will supply 5kw and 14kwh. The EVTV product, slightly more. An EV power set up as discussed above might supply 1/5 -1/10 that power but about the same amount of energy. An imiev or Leaf will run 20+ hours while supplying 500 watts. A Tesla will run a week or longer. ANY EV can supply about 6kwh overnight. That is, an EV can be a significant source of home power and energy when the grid is down. In my scheme, the EV must be feeding into a functional micro grid. The real grid can be used but that would be pointless.
12v battery to grid inverter:
https://tinyurl.com/yxopbw54
EVTV micro grid:
http://store.evtv.me/proddetail.php?prod=Powersafe15
Exercise to grid:
https://courseware.ee.calpoly.edu/~dbraun/srproj.htm
TMC thread on 12v -> grid
https://teslamotorsclub.com/tmc/threads/powerwall-2-0-backup-runtime-extender.126358/
Here is an interesting but not especially practical solution to keeping your grid tie PV power flowing during a grid down situation:
https://insideevs.com/news/324207/getting-your-grid-tied-solar-array-to-work-when-the-grid-is-down/
It is not "success", nor is it "abject failure". I ran one M215 overnight doing about 50 watts. I had been using a clamp on amp meter to estimate how/if things were working. Clamp on amp meters are notoriously inaccurate and noisy. Of course, they do not integrate power into energy. So, I recently installed the instrumentation the TMC thread orginator used. I now measure 12v power and energy out of the car and 240vac power and energy into the grid. Overnight, the ~50 watts gave me: 648 wh from 12vdc, 501 wh to 240vac.
IF I could do that on multiple, say 15-20, M215s, I would consider it a qualified success. But, I am unable to get any significant power out of more than one M125. I lost an estimated 8kwh out of the car battery. Presumably that is mostly unavoidable overhead used within the car when it is "on". As more power is pulled through 12v, the overhead fraction should decline.
"Ray is a genius"
Ray suggested trying powering M215s with my 36v golf cart. I was not optimistic but did the experiment anyway. It was wildly successful. I ran about 230 watts into my house demand most of the night. I started with 32 surplus hoverboard batteries charged to about 38 volts. The 32 is what I normally use for running around the place and I generally run them below 37v before charging. Being surprised with the success I had not thought carefully about how long that battery would run. It ran out at about 2am after having pushed about 2kwh into the house. That is, it relieved the PowerWall of having to supply that energy.
Today, I devoted myself to putting more batteries in/on the golf cart and preparing a 2nd M215 for production. I now have 52 batteries in/on the golf cart charged to about 39.9v. Total battery capacity last night was about 4.6kwh. Tonight, it should be 7.5 kwh.
Efficiency appears VERY good. One M215 pulls about 237 watts from the 36 v battery and delivers about 230 watts into the 240 vac.
Let's see if I can hope to run 2 M215s all night..... they will be pulling about 474 watts for about 12 hours. 5.7 kwh. I don't have the battery fully charged but I believe it has that much in it. So, I'll give it a try. I expect to augment the PowerWall's 14 kwh by an additional 5-6 kwh.
Just for fun, I may try FOUR M215s tomorrow and maybe add some more battery to the golf cart. That will put me close to doubling the PowerWall capacity.
The golf cart battery solution is not as attractive as the 12v from an EV solution I've been working on. There are a lot of EVs around and very few golf carts stuffed with lithium batteries. The 36v batteries can be stationary, not installed in a golf cart, but I do not expect them to be used often to augment the PowerWall or other micro grid. That is, they have the most value when used frequently in a golf cart and occassionally for home power.
9/8/19
I ran three M215s off the golf cart 36v last night. Production was around 700 watts which is about my average night time demand. Again, I ran the battery down about 3:30am. At, I think, about 5.5kwh. That is lower than I expected; I will have to look for trouble with the batteries.
I had a scare. After starting 36v production, I decided to switch off grid. After the switch, the 36v M215s stopped and refused to restart for about an hour. The explanation may be that the PowerWall tries not to allow PV production when the PW battery has a high state of charge; I believe it was at about 90% at that time.
I will remain off grid and observe what happens as PV production rises and the demand of PW and GC charging declines.
7:30 am PV just starting at about 400 watts. Demand about 1500 watts. PW charge state is 47%. Will turn on air conditioner and add about 1500 watts of demand.
8:00 am air conditioner is on. 3.3 kw being pulled from PW battery. PV production about 1kw.
8:30 am. Near equilibrieum. PW battery is charging a bit while AC is cycled off, supplies a bit when AC is cycled on. PowerWall battery at 36%.
9:00 am. The PW is charging at 1-2kw even while the AC is running. PW still at 36%. The GC has been charging at about 800 watts for about 1.5 hours. If I had EV charging to do, this would be the time to start.
9:30 am. The PW charging is threatening to reach the limit of 5kw. If that limit is reached, the PW will shutdown PV production. I am going out to work for an hour or so but will first turn off some of the PV.
10:30 am. About 30 of the panels in the south yard are shut down. Manually via circuit breakers. It turned cloudy and production is varying around about 4kw. The PW is at 49% and is supplying around 1-2kw. I have confidence that the PW will become charged in a few hours. At that time, I will have balancing trouble; the PW will shutdown PV. If I had EVs to charge, my goal would be to keep the PW at 30-50% until early afternoon. Then, assure the PW becomes fully charged as dusk approaches.
11:00 am. Varying sun. PW at 56%. PW is charging one minute, discharging another. Will let things ride for a while. Check on GC charging. Will likely switch on grid soon to avoid squandering the valuable energy that is in danger of being lost. Indeed, since I have ~30 panels switched off, I am squandering energy as I write. The term "switch on the grid" is worrisome. What if the grid were not there to be switched on? Should I have to operate without the grid, I think I would do much as I have this morning. Having an EV or two to charge would a positive. Maybe next time.
1:00 pm. Things went bad sometime after 11am. I came back in the house to find the power had been off. As I observed, the power cycled repeatedly. Down for a few seconds, up for a few minutes. I turned off most of my not roof top PV. Power cycling continued. I turned off the roof top PV, 5kw. Still trouble but I observed low voltage, ~234. Wait for a while. Things stabilized but I was drawing from the PW battery with no PV production. Turn the roof top back on. Things are working fairly well. I'm using about 3kw for air conditioning and GC charging. PW either charges or supplies depending on sun. PW at 55%. I'll let this ride for an hour or so, then turn grid and all PV back on. It looks like I can operate only with the roof top, off grid, in a fairly stable situation. Adding enough PV to only charge a car will likely be troublesome. I'm guessing I might be much better off with two PW batteries which could absorb up to 10kw from PV.
An Overview
I started my PV adventure about 8 years ago with the goal of suppling my own electric energy needs on each of my three electric meters. Especially as PV component prices dropped and with the utilitiy's "net metering" program, that goal was easily, though slowly, met. Then, with attractive rates being offered by my utility, I set a goal of earning as much money as possible from the production of electric energy. That goal is also met though tweaking remains to be done. A much longer term goal is to be able to weather a multi-day grid down event. Installing a PowerWall micro grid at my house meter was a big step in that direction. With the grid there to accept my over production, it is easy to avoid buying ANY power from the utility. That is, when grid connected, my meter always "runs backward", never forward. A major obsticle remains: maintaining PV production when the grid is not there to accept over production. It is a problem of balancing PV production with electric demand.
The next project is to install a micro grid at the big barn / guest house meter so that I can maintain use of both my domestic water pressure and my internet access; both are dependent on that meter's power.
9/15/19
The first 12vdc -> 240vac grid tie inverter arrived yesterday and I tested overnight. Input 14vdc and 250watts. Output 240vac and 211 watts. Disappointing low power and low efficiency. A second unit should arrive soon and I will then test two in parallel. My power source is an imiev and mostly overhead took about half the battery capacity overnight, perhaps 8kwh, to provide about 2khw worth of usable energy. I'll look into pulling fuses to at least get the parking lights off.
9/22/19
My second 12v grid tie inverter arrived yesterday and I tested last night. The 12v inverters are inefficient, pulling 250 watts from 14vdc and supplying only ~200 watts at 240vac. I ran the two from the imiev, pulling ~500 watts. I had pulled the fuse on the imiev which supplies daytime running lights, seemingly a major parasite load. Just for fun, I also ran two M215s from the golf cart. The M215s pull about 237 watts each from the golf cart and supply 230 watts each at 240 vac. Total PW suppliment power was about 877 watts. The golf cart batteries were depleted about 2am; I can add batteries to the golf cart to run three of more M215s all night.
The four grid tie inverters supplied about 7kwh during the over night test. That is a large fraction my night time demand so the PW battery was depleted very little. At times the set up was exporting power to the grid; I could have forced PW battery charging by disconnecting from the grid but I feared what would have happened had the PW battery became fully charged.
Conclusions:
1) ~500 watts is a generous PW supplement and no more is likely to be needed in my situation. A single 230 watt M215 is likely sufficient. In this season, my primary night time demand is a ~600 watt air conditioner; when that air conditioner is not needed at night, the PW easily supplies all needed overnight power. The air conditioner cycles and average power is well below 600watts; the supplements supply power continiously.
2) The 36v golf cart is a workable solution in my situation.
3) The 12v inverters work ok but have objectionable efficiency and their cost is high compared to the more robust and efficient M215s. The much larger EV batteries are an attraction to using the 12v inverters. The high parasitic loads of the imiev are objctionable. I will eventually test with a Tesla and perhaps a neighbor's Leaf.
4) The best solution, IMHO, remains running M215s from 12v through a voltage booster. I have not yet found boosters that work very well. The search continues.
Tesla Model 3 test
The Model 3 puts out only a little more than 12vdc. At that voltage the 12v grid tie inverters supply only about 130 watts. So, about 200 watts with an imiev's 14 volts but only 130 watts with the Tesla's 12 volts.
9/25/19
Several tests with the Tesla 3 (rather than an imiev) reveal that the Tesla is unwilling to give up more than a few hundred watts from the battery. That is, the DC-DC does not try very hard to keep the 12v battery at full voltage. I concluded that the car is monitoring 12v demand and allowing the DC-DC to only meet that demand. Connections to the battery do not get monitored. That was confirmed by a very helpful guy on the TMC thread:
https://teslamotorsclub.com/tmc/threads/powerwall-2-0-backup-runtime-extender.126358/page-5
The bottom line is that the 12v supplemental connection needs to be made under the rear seat rather than at the battery:
https://www.travisllado.com/2019/05/tesla-model-3-stereo-part-9-summary-and.html
I will likely move my 12v tap cable from the front battery to the under rear seat tap point and route the SB50 end of the cable to somewhere behind the rear seats. I'm thinking I should also install my 12v 100 amp circuit breaker.
Though a little more inconvenient for those of us wanting V2H supplemental power, the good news is that Model 3 batteries are likely longer lived. I don't yet know whether S/Xs use the same scheme. Recall that the early S 12v batteries were highly troublesome with, in some cases, a life of only a year.
10/17/19
The 12v tap on the Model 3 IS now under the rear seat. Photos in the TMC post below.
I just made a TMC post which I will not likely repeat here:
https://teslamotorsclub.com/tmc/threads/powerwall-2-0-backup-runtime-extender.126358/page-5
1/2/20
This project has been dormant for several month by virtual of having reached near completition. See recent post:
https://wmckemie.blogspot.com/2019/12/powerwall-agumentation-new-layout.html
More than one day's worth of continious cloudiness prompted me to run some augmenation; the easiest set up is 36v golf cart batteries feeding into two 250w microinverters. The cloudy days prevent all PV energy from being used to charge the PowerWall; at such times, I'm inclined to go off grid to prevent day time power from being sold to the grid. Yesterday, I was off grid all day yet did not get the PW charged beyond 65%. So, I ran the golf cart battery augmentation starting about 4pm. The starting charge was about 39 volts. Though the max charge voltage is 42v, I generally do not charge beyond 40-41v in an effort to prolong battery life. This morning, after PV started producing, my golf cart batteries were down to 25v, nearly exhausted. At that time, the PW battery was down to about 19% and could have gone another several hours. Looking at my Curb data, I estimate that the golf cart batteries supplied about 6kwh. So, the total available, starting from a near full charge, would be about 7kwh. That is a total of about 50 hoverboard batteries with a capacity of about 4ah (about .14kwh) each.
No comments:
Post a Comment