Mitigation of demand from utility with solar

[phx]

SB5000US-11 is a pretty old model manufactured in 2012. In the end of last year (when I submitted application for utility permit) it was still on their approved list. I was lucky that my system was approved and commissioned before utility removed that model from the list of approved equipment. So I purchased it in mint condition from Ebay for very decent price

It has GFDI. It looks pretty same as GFCI. I believe as long as I keep NEC-compliant grounding of inverter enclosure and negative DC line according to specifications, it won't interfere

Hedges raised a good concern about how to get 338V from battery inverter if neutral in AC side of it is also bonded to ground. AFAIK all those "true sine" inverters all use something around 50kHZ on MOSFET's and ferrite transformer in their internal circuitry. Then they rectify it to DC and use second DC-to-60Hz high voltage inverter in the same box to produce 60Hz rectangle or "true sine" output. In latter case I guess they fiddle with PWM either between rectifier and second inverter or maybe with MOSFET's before transformer to form that "true sine". Using PWM is the only way how MOSFETs can avoid working in analog mode which would instantly kill them with overheating

In any case they should have a 50kHz rectifier unless they have a heavy and bulky 60Hz direct upconvert transformer. So for my purpose I can grab 338V DC directly after that internal 50kHZ rectifier within a box and disconnect rectifier from any further 60Hz-producing circuitry thus insulating it from the ground.

Also benefits of 50kHz rectifier are much smaller capacitors than needed to rectify 60Hz on inverter's conventional output

I searched alibaba.com. Some vendors there have step-up convertors from 24V DC to 310V DC or 48V to 310v or 480V. I don't remember the exact output voltage. But it's within normal operational limit of SB5000US-11. So that DC-DC convertor might be the best choince for my case. The only caveat, it doesn't have a built-in BMS

I don't trust high voltage batteries. They are very expensive and need expensive BMS. Also they are not very common and if one cell fails, entire battery is wasted. Hence they are not reliable. All those reasons look like "Show stopper" from using "Sunny Boy Storage" requiring high voltage battery

 

[Hedges]

You just beat the "Rule 21" grid support which obsoleted the older models.

Large UPS have series strings of gel or AGM batteries. Ideally a lithium battery would be configured so you could isolate and bypass a failed cell, which is only 1% of the high voltage pack. But not necessarily easy for DIY. Maybe a high-voltage vehicle battery. But anything battery direct to Sunny Boy would need a soft-start charging capacitors gradually. I figured that can be done through a resistor starting from zero because Sunny Boy waits 5 minutes to evaluate grid before producing power. But a Sunny Boy which is already producing from PV would immediately start drawing current through your resistor when battery connected and you would have a larger thermal load to deal with. Not impossible, just more engineering.

Yes, a suitable boost switcher should do it. Maybe add an over-voltage crowbar to protect Sunny Boy, because output voltage of boost rises uncontrolled unless active regulator functions correctly.

Switchers either use a fast diode, or "synchronous rectifier", the MOSFET itself for lower voltage drop. For high voltage output a diode represents small percentage.

Motor-generator would be another (less efficient) way. But more chance of over-voltage; protection circuits are used between wind turbine an inverter.

 

[ggeorge063]

Newbie here, curious about installing solar PVs for my home. Just have a few questions for you. How has your investment in your personal solar system reflected in your bills so far? I'm curious how it fares in the summer since it's hot, there's a lot of sunlight, and the average cost of electricity bills in Phoenix (read here) tend to fluctuate due to increased usage from A/Cs. I'm looking to get into solar systems for my home as well and I'm a frugal guy so I'd like to make sure my investments becomes ROI in the long run. I have no past experience with these kinds of things and I'm trying to gather all the information I can from first-hand sources so I can make a wise decision on how to work with this, not just based off paid articles or marketing gimmicks etc. TIA.

 

[Hedges]

I put in PV 15+ years ago, guessing incorrectly that PV costs would not go down, up if anything (PV takes area, doesn't follow the performance curve of Moore's law. Energy is a substantial component of PV cells and aluminum frames. Dollar was getting killed.)
After 50% CEC rebates and amortized over 15 years, PV has cost me $0.20/kWh which is break-even with utility.

Today I can easily put in grid-tie PV for $1/watt or $0.05/kWh amortized over 10 years. Inverter and PV panels is half of that, so generously allowing balance of system to cost as much. Being frugal (cheap, use scrap lumber for mounts, surplus electrical) I could make power for $0.025/kWh. After a decade, possibly replace inverter if failed, and enjoy power for $0.01 or the next decade.

Grid-tie with net metering is the way to go. The Gotcha! is if you are forced onto a time-of-use rate schedule where you pay $0.50/kWh during late afternoons when you have to run A/C, but are credited $0.05 earlier in the day when you have PV production. Caveat Utility Customer.

By putting in sufficient grid-tie inverter capacity, and over-paneling to make up for reduced output during hot conditions and late in the day, you should be able to be a net producer rather than consumer up to 6:00 or 7:00 PM in the summer. Maybe be net zero between 4:00 and 9:00 PM, which is my utility's peak time off use. In that case you could do OK. The utility will change the rules of the game as time goes on.

 

[phx]

Newbie here, curious about installing solar PVs for my home. Just have a few questions for you. How has your investment in your personal solar system reflected in your bills so far? I'm curious how it fares in the summer since it's hot, there's a lot of sunlight, and the average cost of electricity bills in Phoenix (read here) tend to fluctuate due to increased usage from A/Cs. I'm looking to get into solar systems for my home as well and I'm a frugal guy so I'd like to make sure my investments becomes ROI in the long run. I have no past experience with these kinds of things and I'm trying to gather all the information I can from first-hand sources so I can make a wise decision on how to work with this, not just based off paid articles or marketing gimmicks etc. TIA.

This is my energy report for almost entire year in Phoenix, AZ. Please see a screenshot

My system was commissioned with just 42 panels and one inverter 7.7kW in November, 2019. Then I upgraded it with second inverter to 12.96kW and 64 panels total in April. It covers 93% average of 11 months electricity consumption

Considering that January-March it was running with smaller power, I would estimate that it covers more than 100% of my energy consumption per year

Typical A/C run in Phoenix in July-August is about 14-18 hours a day. But solar production covers 2/3 of it. For the rest of months it covers even more than 100%

However demand charges for peaks circled with red seriously impact benefits from solar. Nevertheless even with harsh demand rip-off fees I still pay about half of what I paid without solar per month

 

[Hedges]

Could you store energy by building ice off-peak, then blow cool air (or circulate antifreeze mixture) to chill the house on-peak?

H2O is a lot cheaper than Li or Pb. Or demand fees.

 

[phx]

Could you store energy by building ice off-peak, then blow cool air (or circulate antifreeze mixture) to chill the house on-peak?

H2O is a lot cheaper than Li or Pb. Or demand fees.

That's pretty interesting idea. But if I understand correctly, 1Ton in HVAC is amount of energy needed to melt 1 weight ton of ice at 32F degrees within 24 hours. Please correct me if I'm wrong since I'm not a HVAC specialist. So if A/C is 5Ton, it means it can melt 5 ton of ice in 24 hours of constant running. Considering most critical part of peak time in summer is 5-8pm (i.e. 3 hours), that would require ice tank with capacity (5*3)/24 ton (about 170 gallons) of water to freeze it to ice and back

That's a pretty huge tank. So several questions instantly come:

• What material can withstand freezing of 170 gallons of water? Metal will be torn by ice at freezing
• How to protect such tank from 110-120F Arizona heat? THe only possible solution might be underground storage and multiple layers of styrofoam or fiber insulation around tank
• All piping system including radiators to convert existing HVAC system from air-based to ice-based
• Ice making machines are not so common as household A/C systems. So It's hard to find one for reasonable price and build a working system for described goal

Any further ideas at least for the proof of concept?

 

[Hedges]

About 8 gallons per cubic foot, so 20 cubic feet would be 160 gallons. 2'x 2' x 5', not so huge. Like a large commercial water tank.
Have a crawl space with room?

Water expands when freezing to ice. Maybe a bladder inside compressible foam insulation. (or no bladder, just bond foam to bottom and sides of tank.)

Consider a propylene glycol loop in tubes to extract cold, and going through a radiator to chill HVAC air.
I have an electric duct heater downstream of my furnace, and a switch to select gas heat vs. electric. Yours would select A/C compressor vs. circulator pump.

Extracting cold is easy. As glycol loop melts ice, convection transfers heat from loop to remaining ice.
Freezing is slower, because ice stops convection and provides some degree of thermal insulation.
But more time to cool - my off-peak hours are Midnight to 3:00 PM, 15 hours.
Perhaps the refrigeration system (separate from A/C) would have expansion valve and copper loop in water. Longer than glycol loop due to lack of convection.

These systems are installed commercially, question is how to build cost-effectively.
My first though was just using a chest freezer. But probably insufficient cooling capacity - 300W vs. 3000W for A/C of my small house.

Another implementation builds ice at night (when more efficient than dumping to hot air during the day), and uses ice rather than air for condenser of the A/C which runs during the hot day. More efficient, so less power consumed during peak.

Ice storage air conditioning - Wikipedia

en.wikipedia.org

ICE-PAK® Ice Storage Units

Enjoy optimum performance and compact use of space.

www.evapco.com

 

[ggeorge063]

This is my energy report for almost entire year in Phoenix, AZ. Please see a screenshot

My system was commissioned with just 42 panels and one inverter 7.7kW in November, 2019. Then I upgraded it with second inverter to 12.96kW and 64 panels total in April. It covers 93% average of 11 months electricity consumption

Considering that January-March it was running with smaller power, I would estimate that it covers more than 100% of my energy consumption per year

Typical A/C run in Phoenix in July-August is about 14-18 hours a day. But solar production covers 2/3 of it. For the rest of months it covers even more than 100%

However demand charges for peaks circled with red seriously impact benefits from solar. Nevertheless even with harsh demand rip-off fees I still pay about half of what I paid without solar per month


View attachment 28975

Just what I was looking for, thanks man!

 

[wmills_92105]

Sounds like your options are to add battery into you main setup and make a ton of changes or set up a kind of reverse setup on dedicated circuits for the heavy load AC unit.

• A transfer switch that prefers your new inverter to utility power.
• Inverter and battery bank big enough for you load that has an on off control cir unit for low voltage that you can reuse for timing control
• Timer to control your inverter
• Timer to control a probably separate charging system. Maybe pull the same trick if you get a charger that accepts a overvoltage signal wire and drive with a timer. Or just a straight timer on the input power.
• Battery system has to protect itself from over/under voltage.

Designing on the fly by an amateur should be considered carefully by experts...

 

[whittakerj]

Try this.

Hacking your SRP time-of-use bill with a DIY Tesla Powerwall made from recycled scooter batteries

I’ve always been fascinated by technology. I love building all types of electronic projects and automating things. Recently I ran across a guy who discussed building a battery backup using recycled scooter batteries that contained 18650 rechargeable batteries. The same … Continue reading →

jeremywhittaker.com

 

[Ampster]

Try this.

Nice graphics and analysis but why Scooter batteries? Why not LFPs like many of the posters on this forum. Some are seeing costs as low as $0.12 per Watthour.

 

[whittakerj]

It was just a project I did last year. It was simple and cheap. You can mitigate 1kwh with 2 batteries. In the original post he needed to get rid of 8kwh. This could be done with about $800 of those batteries. They use the Lithium-Ion 18650 batteries, same as Tesla.

 

[whittakerj]

Nice graphics and analysis but why Scooter batteries? Why not LFPs like many of the posters on this forum. Some are seeing costs as low as $0.12 per Watthour.

I'm not familiar with LFP at $.12 per watt-hour. Where can these be purchased?

 

[Ampster]

Where can these be purchased?

Welcome to the forum. Lots of resources here for those. Start with this thread:

Introducing the New & Improved Shenzhen Luyuan Technology Co.

You need to go after Fed Ex from what I see. Put in a claim. From what I’ve understood, claims to FedEx can only be submitted by the shipping party...

diysolarforum.com

 

[740GLE]

I do not understand those $17.51 for each kW below 10 and $33.59 for each kW above 10.

That is the most ridiculous rates ever. In the USA no less and in Phoenix Arizona?

Is anyone else in the USA at this forum paying such high rates?
I am curious to know if I am really fortunate (for the first time in my life) to live in California.

Is it to punish those on solar so they pay more than if they were on grid?
What is the normal Joe on grid KW-hr rate?

I think because I am an energy hog at 15KW-hr per day I am in the incremental $1 per KW-hr for those last few KW each day.
At $30 per KW-hr X 15Kw, I would be paying $450 per day and $13,500 per month!

Are you sure about those rates? That is exorbitant and something I have never heard of.

If those rates are true I would ditch the solar and get in a normal rate plan and save tons of money!

Then secretly install one of those on grid solars to supplement your electricity.

He prob confused $ and ¢. It prob takes into account with possible net metering policies” when the grid is used as a “battery” when a system over generates.

Remember, there’s no free lunch. If solar generation falls off with the sun, the utility still needs to keep the lights on no matter how much a solar system generated at noon, someone’s gotta pay for that make up.

Large scale customers (commercial industrial customers) are charged similar fashion, they grid is sized/designed for the largest demand, if the grid needs to be updated to meet larger demand, demand fees account for that.

Just think of all those Tesla super chargers rated at 100-150kw, plug 5-10 cars in to charge at once, that’s a huge demand, Tesla pays the fees to support the demand.

 

[phx]

Unfortunately there is no confusion between dollars and cents per kW when SRP calculates demand fees. Please see my posing in this thread from
Jun 27, 2020 with explanations of those ridiculous rip-off charges

 

[whittakerj]

I do not understand those $17.51 for each kW below 10 and $33.59 for each kW above 10.

That is the most ridiculous rates ever. In the USA no less and in Phoenix Arizona?

Is anyone else in the USA at this forum paying such high rates?
I am curious to know if I am really fortunate (for the first time in my life) to live in California.

Is it to punish those on solar so they pay more than if they were on grid?
What is the normal Joe on grid KW-hr rate?

I think because I am an energy hog at 15KW-hr per day I am in the incremental $1 per KW-hr for those last few KW each day.
At $30 per KW-hr X 15Kw, I would be paying $450 per day and $13,500 per month!

Are you sure about those rates? That is exorbitant and something I have never heard of.

If those rates are true I would ditch the solar and get in a normal rate plan and save tons of money!

Then secretly install one of those on grid solars to supplement your electricity.

It is a one-time charge. Meaning they take the highest kWh usage in a 30-minute block of time from the hours of 2pm-8pm for the entire month. Then they apply those multipliers.

More details can be found here - https://www.srpnet.com/prices/home/residentialdemand.aspx

 

[phx]

Here is demand chart for last 30 days. SRP will charge for 7.1kW peak at 8pm on 4/29/21 ignoring all my efforts to keep demand below 5kW for the rest of the month.

What a shame with such rip-off for 20 minutes running laundry dryer and oven at the same time! That even considering that A/C was already turned off by my DIY "Smart Demand" system described above in this thread.
?

 

[740GLE]

Ahh i missed the scaling flat rate fee, vs a scaling kwhr rate.

So just modify your habits and save the coin? I run my dishwasher on a 4hr delay, laundry when we head to bed.

Why can’t you shut off the AC system after you lose peak solar, then turn it back on? If your house heats up in 3 hours of no AC sounds like you have an insulation/weatherizarion issue which leads to your crazy high usage.