Thinking things through/Planning an endpoint

[KevinC_63559]

A wise man once told me that its easier to get someplace if you know where your going. Realizing I'm playing around buying bits and pieces without really having thought out my end-point, or rather, having thought out several end-points and not picking one.

So my current train of thought, in priority order:

1) Need to be able to provide power to my fridge and freezers during outages similar to the 37 hour one we just lived through. Key is that these devices only need power for a few hours a day, but also that most long outages occur during snow storms, with half-day outages due to wind occasionally. Current solution is to use an 1800W portable gas powered backup generator with extension cords - presuming it starts and presuming I was wise enough to stock up on gas.

2) Want to be able to replace three UPSes. They do a reasonable job on short outages, but lost month-old battery packs during this 37 hour one. Alas, they are based on 12V9A batteries, with inherent run time limitations, and don't have charging circuits designed for LiFePO4 batteries. If I just wired, say, all of my 120V stuff to the future battery/inverter system, they become redundant.

3) Nice to have some 12V power for a SDR setup, and similar 12V (all of this refers to my man cave building). 10amps of 12V would be plenty. Would it be legit to just tap across one 12V battery, even if that battery was in series. Feels wrong, like that battery would get out of balance., but maybe not - especially is a full 24 (48?) volt BMS was used vs. individual 12V systems. Suppose I could just rig one small 12V system separate from the freezer one.

4) Be great to be able to peak shave in order to get some daily usage out of the system.

Odd to say, but good news is that my electric company has just started charging everyone a Peak Demand fee. I'll see that bill in a few days. My peak demand usage would be, I suspect, a great indicator for the size of inverter needed. Once an inverter size is determined, I can make a call on 12/24/48 voltage question.

Random garbage floating in my head:

A) Under the impression the end-state battery arrangement needs all new batteries. e.g. Its a bad idea to run banks with mixes of new and not-so-new batteries. Got that impression back in Pb battery days, but suspect it holds for LiFePO4 chemistry as well, although perhaps the BMSes could compensate? If this is true, the DIY battery built from individual cells, partly justified by being able to replace a single bad cell, is a fallacy, ?right? (since all cells should be of a similar age)?

B) Like the idea of a pass-through inverter, like the Victron Energy units, for general usage. Grid power is cheaper than solar, but its still not cheap. Seems a shame to lose efficiency via an always on inverter. Unsure, and feels a bit redundant, if I could use solar MPPT charges to augment power to a Victron based battery bank. Perhaps if the Victron could be told to only charge is the batteries were below, say, 30%, giving my default the MPPT chargers priority?

C) If I size a solar array to run the load for a few hours a day (point 1), in the winter, on cloudy days, that array is going to be wasted most of the year. So peak shaving (point 4)? or load sharing (say run off the batteries if they are > 60% charged, and only grid charge if below 30%)?

D) What am I looking for? Presuming a hybrid inverter based system? Again, thinking the inverter selection is going to drive most everything else. Recommended brands?

E) Presuming a hybrid inverter, most of which I've seen have their own MPPT charge controller, does this eliminate me being able to use the DIY Solar For U 20 amp MPPT charge controllers (presuming I end up with a 12 or 24V system)?

Brain is starting to hurt... comments/questions/clarifications/suggestions all welcome.

Thanks!

 

[sunshine_eggo]

1) Need to be able to provide power to my fridge and freezers during outages similar to the 37 hour one we just lived through. Key is that these devices only need power for a few hours a day, but also that most long outages occur during snow storms, with half-day outages due to wind occasionally. Current solution is to use an 1800W portable gas powered backup generator with extension cords - presuming it starts and presuming I was wise enough to stock up on gas.

Whether they only need a few hours a day, or you leave them plugged in continuously, they're going to draw about the same amount of total energy every 24 hours.

How much energy do they use in a 24 hour period? Do they have a yellow energy guide sticker, or have you placed a kill-a-watt or similar meter on it?

2) Want to be able to replace three UPSes. They do a reasonable job on short outages, but lost month-old battery packs during this 37 hour one. Alas, they are based on 12V9A batteries, with inherent run time limitations, and don't have charging circuits designed for LiFePO4 batteries. If I just wired, say, all of my 120V stuff to the future battery/inverter system, they become redundant.

You may want to retain these even with a backup system in place. Many lower cost inverters don't switch over fast enough to prevent disruption. The only solution for cheap inverters is to have them powering the load continuously from battery with a grid powered charger keeping them topped off.

3) Nice to have some 12V power for a SDR setup, and similar 12V (all of this refers to my man cave building). 10amps of 12V would be plenty. Would it be legit to just tap across one 12V battery, even if that battery was in series.

For the love of all that is holy, that's among the worst things you can do. NEVER do it. EVER.

Feels wrong, like that battery would get out of balance., but maybe not - especially is a full 24 (48?) volt BMS was used vs. individual 12V systems. Suppose I could just rig one small 12V system separate from the freezer one.

Will do so in the extreme. Will bring your 24/48V system to its knees in short order.

4) Be great to be able to peak shave in order to get some daily usage out of the system.

Depends on the design and features of the equipment. The aforementioned AC charger option would work if you simply had an AC timer on the charger such that it only draws on off-peak time.

A) Under the impression the end-state battery arrangement needs all new batteries. e.g. Its a bad idea to run banks with mixes of new and not-so-new batteries. Got that impression back in Pb battery days, but suspect it holds for LiFePO4 chemistry as well, although perhaps the BMSes could compensate? If this is true, the DIY battery built from individual cells, partly justified by being able to replace a single bad cell, is a fallacy, ?right? (since all cells should be of a similar age)?

Within reason. Batteries in a string need to be as identical as possible. Parallel strings can be notably different.

B) Like the idea of a pass-through inverter, like the Victron Energy units, for general usage. Grid power is cheaper than solar, but its still not cheap. Seems a shame to lose efficiency via an always on inverter. Unsure, and feels a bit redundant, if I could use solar MPPT charges to augment power to a Victron based battery bank. Perhaps if the Victron could be told to only charge is the batteries were below, say, 30%, giving my default the MPPT chargers priority?

Something like the Multiplus II has a maximum idle draw of 11W. There are a variety of ways to charge based on SoC and to give non AC chargers priority.

E) Presuming a hybrid inverter, most of which I've seen have their own MPPT charge controller, does this eliminate me being able to use the DIY Solar For U 20 amp MPPT charge controllers (presuming I end up with a 12 or 24V system)?

Hybrid generally includes the potential for grid feedback. You likely don't want this as it's an added layer of complexity and regulations.

I'm not a big fan of All-in-ones (AiO) that have all components in one box. One goes bad, you may lose all functions.

I'm still perceiving you as being on step 1 that I recommended in the other thread, but you are very distracted.

 

[KevinC_63559]

Whether they only need a few hours a day, or you leave them plugged in continuously, they're going to draw about the same amount of total energy every 24 hours.

How much energy do they use in a 24 hour period? Do they have a yellow energy guide sticker, or have you placed a kill-a-watt or similar meter on it?

Put one on today. I know they draw about 800W when first plugged in, but that drops rapidly to 350W, then 80W, currently at 0W. Obviously if they were offline for awhile, they would draw the max until things were rechilled to a reasonable level. After that it would depend on thermal drift in the container.

You may want to retain these even with a backup system in place. Many lower cost inverters don't switch over fast enough to prevent disruption. The only solution for cheap inverters is to have them powering the load continuously from battery with a grid powered charger keeping them topped off.

Victron claims under 20ms, so a bit over 1 60Hz cycle. Victron specification sheets bother me though, often stating they produce 50 Hz power, so that 20ms would be 1 cycle. I presume everything they sell stateside is 60Hz.

For the love of all that is holy, that's among the worst things you can do. NEVER do it. EVER.

Thought banished.

Depends on the design and features of the equipment. The aforementioned AC charger option would work if you simply had an AC timer on the charger such that it only draws on off-peak time.

Rural power coop. Peak Demand is this decades change. Maybe off-peaking charging next decade?

Within reason. Batteries in a string need to be as identical as possible. Parallel strings can be notably different.

Good to know. Thanks.

Something like the Multiplus II has a maximum idle draw of 11W. There are a variety of ways to charge based on SoC and to give non AC chargers priority.

Any recommended reading?

Hybrid generally includes the potential for grid feedback. You likely don't want this as it's an added layer of complexity and regulations.

Yeah... and our co-op is not friendly to grid feedback, charging something like $120/month for a bi-directional meter (vs. $37/month for a regular 200amp meter). If/when I get to the house, it will be my goal to use the grid as a source of cheap backup power - topping off batteries at night, that type of thing.

I'm not a big fan of All-in-ones (AiO) that have all components in one box. One goes bad, you may lose all functions.

Excellent point. Tis why I always build deskside computers instead of using laptops - I can always tweak a tower. Laptops are pretty much what you get. That was one of the things I liked about the DIY Solar For U MPPT charger - at 20amps, 700 Watts (@24V), it felt like a nice "chunk" size. Having a bit of a problem maxing that design out with modern 360W panels, but that is another topic.

I'm still perceiving you as being on step 1 that I recommended in the other thread, but you are very distracted.

A lot to learn. But I'll get there. Just considering as much as my brain can handle for now, perhaps a bit more <smile> Appreciate all your feedback. Good/Fast/Cheap - pick two was an IT motto I lived by for decades. Finding a balance point one is happy with is the challenge.

 

[sunshine_eggo]

Put one on today. I know they draw about 800W when first plugged in, but that drops rapidly to 350W, then 80W, currently at 0W. Obviously if they were offline for awhile, they would draw the max until things were rechilled to a reasonable level. After that it would depend on thermal drift in the container.

My guess is that's probably pickup up surge and run current. The thermostat draw is probably below the detectable threshold.

However, you run them, they're probably going to use ~1.5kWh/day - ish.

Victron claims under 20ms, so a bit over 1 60Hz cycle. Victron specification sheets bother me though, often stating they produce 50 Hz power, so that 20ms would be 1 cycle. I presume everything they sell stateside is 60Hz.

They have 230/50Hz and 120/60Hz product lines. Their 230/50Hz can run at 240/60Hz and be combined with an autotransformer for 120/240VAC split phase.

Any recommended reading?

Yeah... and our co-op is not friendly to grid feedback, charging something like $120/month for a bi-directional meter (vs. $37/month for a regular 200amp meter). If/when I get to the house, it will be my goal to use the grid as a source of cheap backup power - topping off batteries at night, that type of thing.

OUCH!

Excellent point. Tis why I always build deskside computers instead of using laptops - I can always tweak a tower. Laptops are pretty much what you get.

Same. I've been upgrading the "same" tower since 1990. Obviously nothing is the same as it was back then, but I've never bought a whole computer at one time except back in 1990.

That was one of the things I liked about the DIY Solar For U MPPT charger - at 20amps, 700 Watts (@24V), it felt like a nice "chunk" size. Having a bit of a problem maxing that design out with modern 360W panels, but that is another topic.

Problem is it's a very limited chunk size, and it's very expensive for what you get.

A lot to learn. But I'll get there. Just considering as much as my brain can handle for now, perhaps a bit more <smile> Appreciate all your feedback. Good/Fast/Cheap - pick two was an IT motto I lived by for decades.

Aviation & manufacturing too.

Finding a balance point one is happy with is the challenge.

That's why building a robust plan with the final design up front is important. You can really break it down into chunks and avoid unnecessary expenses.

 

[KevinC_63559]

My guess is that's probably pickup up surge and run current. The thermostat draw is probably below the detectable threshold.

Agreed.

Brief mention in here that the VIctron added SoC based source selection in the patch. Cool.

Problem is it's a very limited chunk size, and it's very expensive for what you get.

Currently have a Victron SmartSolar 150/45 on my list. Awaiting feedback from Victron on FCC class B compliance. At around $200, it support 12/24/48V which is convenient (buy its once, evolve the system over time). Of course, awaiting my energy audit results before I can call it my "chunk" device, but the specs seem to line up with with a 2x2 360W STC array (for 24V usage). FWIW: I like that their specs are nominal, not STC.

Hmmm, please sanity check my math: The panels I'm looking at at 360W STC, 272W NOTC, 11.31 Short Circuit Amps, 49.8V Open Circuit. So a 2x2 array would generate 544W @ 24V with 22.62 amps at Short Circuit, and 99.6V Open Circuit right? The 150/45 is rated at:

650W@12V, 1300W@24 Nominal PV power, with controller input power limiting. So I should be good for twice as many panels (handy since I'll probably buy a 10 pack). In fact, looks like this would be compatible with even higher power panels, giving me some selection options there.

50A Max. PV Short circuit current - CHECK

So far, so good?

How much will I be losing by not have a buck/boost charge controller? Is this something I should worry about given that panels have become one of the cheaper components in the system?

That's why building a robust plan with the final design up front is important. You can really break it down into chunks and avoid unnecessary expenses.

Working on it... one piece at a time. OK, that's a lie... trashing over all the components, but feels like things will align eventually.

 

[Steve777]

Just a thought, when you are adding up watt-hours, any auto defrost (aka frost free) refrig or freezer has a "defrost cycle" which runs on a timer. That defrost cycle turns off the compressor and turns on a heating element to melt the accumulated frost/ice and let the melt water drip down to a pan at the bottom under the unit. How often and how long it runs and and how big the heating element is varies by unit.

But if you just measure the compressor power usage for a bit of time and extrapolate from there, you likely will not have caught the defrost cycle and may be under in your power calculations.

 

[sunshine_eggo]

Agreed.

Brief mention in here that the VIctron added SoC based source selection in the patch. Cool.

That's been an option for quite awhile. I've been using my DIY battery since 7/2022, and it's been using the Batrium BMS for SoC.

Currently have a Victron SmartSolar 150/45 on my list. Awaiting feedback from Victron on FCC class B compliance. At around $200, it support 12/24/48V which is convenient (buy its once, evolve the system over time). Of course, awaiting my energy audit results before I can call it my "chunk" device, but the specs seem to line up with with a 2x2 360W STC array (for 24V usage). FWIW: I like that their specs are nominal, not STC.

That's the most budget friendly 150Voc/48V MPPT for sure. I have a 250/100. It hurt.

1. Safety precautions

www.victronenergy.com

1.5. SmartSolar MPPT 150/45 FCC and Industry Canada Compliance​

This device complies with part 15 of the FCC Rules and to RSS of Industry Canada.

Operation is subject to the following two conditions:

1. This device may not cause harmful interference, and
2. This device must accept any interference received, including interference that may cause undesired operation.

Hmmm, please sanity check my math: The panels I'm looking at at 360W STC, 272W NOTC, 11.31 Short Circuit Amps, 49.8V Open Circuit. So a 2x2 array would generate 544W @ 24V with 22.62 amps at Short Circuit, and 99.6V Open Circuit right? The 150/45 is rated at:

Yep.

650W@12V, 1300W@24 Nominal PV power, with controller input power limiting. So I should be good for twice as many panels (handy since I'll probably buy a 10 pack). In fact, looks like this would be compatible with even higher power panels, giving me some selection options there.

50A Max. PV Short circuit current - CHECK

So far, so good?

Yep. 2S4P of those panels meets all requirements. if you were to split your array into different orientations, where it's not possible to produce >50A Isc, you could go even more. 2S6P comprised of 3X separate 2S2P arrays facing E, S and W would still meet the 50A requirement since we only have one sun.

AND the 50A limit exists because the MPPT will protect itself from reverse polarity by shorting the PV circuit. That current is the maximum current the latching circuit can handle. You could also protect the input with a fuse. If you don't hook it up backwards, there's no hard limit... just a practical one... like why would you want to put 50kW on a controller only capable of 2600W output?

How much will I be losing by not have a buck/boost charge controller? Is this something I should worry about given that panels have become one of the cheaper components in the system?

Boost controllers are niche and almost never needed. The benefit of MPPT is higher PV voltage for lower losses with longer wire runs.

Working on it... one piece at a time. OK, that's a lie... trashing over all the components, but feels like things will align eventually.

You'll win of you don't spend the $ before its time.

 

[KevinC_63559]

Just a thought, when you are adding up watt-hours, any auto defrost (aka frost free) refrig or freezer has a "defrost cycle" which runs on a timer. That defrost cycle turns off the compressor and turns on a heating element to melt the accumulated frost/ice and let the melt water drip down to a pan at the bottom under the unit. How often and how long it runs and and how big the heating element is varies by unit.

But if you just measure the compressor power usage for a bit of time and extrapolate from there, you likely will not have caught the defrost cycle and may be under in your power calculations.

Appreciate that Steve, but none of my refrigeration has auto-defrost. That's on me to do every year or so manually. The concept of warming a surface inside a cold box just to evaporate frost off always seemed just, well, wrong to me <smile>. For similar reasons I only use chest freezers for efficiency (beyond the little one associated with our fridge). We are a small cattle farms, so typically have a 26 cubic foot chest freezer (sometimes two, seldom three), full of beef. This should be of no surprise, but: Freezers that auto-defrost tend to have more freezer burn on items than those that don't.

 

[KevinC_63559]

Hmmm... this mornings rumblings (ok, afternoon now): Combine points 2 & 3 utilizing a Victron Multiplus Compact (12V) and solve the 12V source and UPS problems. Since the Compact unit uses 120V input, I can presumably just plug it into a 120V 20amp socket? I'm under the impression the Multiplus can limit the "shore" amperage down to something like 8.4amps. Presumably I could tweak that up a bit (wouldn't have a dedicated circuit, so the legal limit of 16amps would be unwise, but perhaps a nice round number like 10amps). Output to a good surge protected power strip and the UPS issue is solved. Tap of the 12V bus (if I eventually end up with more than one battery) for the SDR related needs. Whole thing would comfortably fit under my desk where the existing UPSes are.

This would let me use my current 12V100A battery and give me some experience with the Multiplus operation.

I'm kind of inverting the sequence of things... but start as above, then invest in a 48V Multiplus (maybe a II) and start rewiring my panels, still just using grid power plus a battery array. Then add solar panels and chargers later. Basically provide my whole office with UPS capabilities. Figure out along the way if I want to invest in Victron Cerbo and GX Tough devices of if the basic systems do what I want.

Bonus: When I start doing the 48V system, I could start with (4) 100A batteries for cost control, then migrate those to the 12V unit for extra capacity (presuming that mixing strings of 100amp and 285amp is not a good idea, or that I someday want to max out at 8 strings of 285amp batteries).

Thoughts?