Generator install - where to start

[WinterDeere]

Good advice, as always, Peter! But do note that my 35 kW number is steady state, without considering surge/inrush of each appliance, which could easily be 100 kW for a brief few seconds immediately following a long outage without load management.

Simply put, I summed the steady-state draw of each appliance, in calculating that load. Add maybe 4 kW for lighting, computers, TV's, refrigerators... we're likely at 35 kW max steady state, if all HVAC's, pool heater, well pump and shop air compressor run at the same time. Heck, now that I think of it, I didn't even include kitchen range (10 kW?) or microwave ovens (2x 1.5 kW?). Now we're at 45 kW.

The scenario of all AC compressors, refrigerators, pool heater, and air compressor running simultaneously for minutes or an hour is very likely, immediately following any extended outage. Although with our goal being the avoidance of extended outages, I guess all of those large loads may never run concurrently under normal circumstances.

 

[Larry Caldwell]

Nothing surprises me anymore - we asked for a bid to install a propane line from our LP tank to the generator last fall.

When I called them, they said, oh, we factored in digging the trench!!

Hi Nicole

Ryan came by today promptly at 8:30 am then called back about 1 pm with the estimate $10,055 - yes $10 thousand dollars to run and connect 10’ of gas pipe and fittings for a 10 kw generator in a trench we will dig/provide and on a pad and post we provide.

Ryan was very professional and pleasant and we were pleased with his assessment and overall presentation as a professional.

However we will not be engaging with your company.

For small projects I rarely ask for a bid. Mostly I go cost-plus. Less paperwork for the contractor, less oops pricing for me. I got a new 200 amp house electrical panel for $780 that way, including permits and a couple hours of electrician time to install it.

 

[Larry Caldwell]

Good advice, as always, Peter! But do note that my 35 kW number is steady state, without considering surge/inrush of each appliance, which could easily be 100 kW for a brief few seconds immediately following a long outage without load management.

Simply put, I summed the steady-state draw of each appliance, in calculating that load. Add maybe 4 kW for lighting, computers, TV's, refrigerators... we're likely at 35 kW max steady state, if all HVAC's, pool heater, well pump and shop air compressor run at the same time. Heck, now that I think of it, I didn't even include kitchen range (10 kW?) or microwave ovens (2x 1.5 kW?). Now we're at 45 kW.

The scenario of all AC compressors, refrigerators, pool heater, and air compressor running simultaneously for minutes or an hour is very likely, immediately following any extended outage. Although with our goal being the avoidance of extended outages, I guess all of those large loads may never run concurrently under normal circumstances.

145 amps? I don't think so.

 

[WinterDeere]

145 amps? I don't think so.

Yeah... there's the rub. I guess the tech just hasn't caught up with reality, yet. Every house in my neighborhood has at least 400A supply, I'm just lucky I have an older house that's still run on oil and wood, rather than electric heat.

And note in that last post, with kitchen range and microwave ovens, we could peak 187 amps. Not that I can imagine a scenario where I'd be intentionally running two microwaves, a kitchen range, and my shop's air compressor. It becomes a game of odds, what combinations of major appliances might actually run together at any time, with the biggest risk probably only occurring in the first minutes following an extended outage.

Because our peak demand only occurs in summer, and neither my own monitors nor the electric company store "by the second" data that long, I don't really have the much data available to see a realistic range of loads for a wide range of dates in mid-July. All I can do is add up all of the steady-state draws, and assume there's some possibility they might occur together.

However, I do have one 3 hour window of data parsed by the second, from 10 July 2024, roughly noon - 3pm. The maxium draw I see in that 3-hour window is only 16.5 kW, but I'd be willing to bet the pool heater wasn't running, and I did not have the AC cranked in the shop or studio at the time, nor was I probably working on something that had the air compressor running.

I also have a week's worth of history stored from the same date, but averaged per minute. The biggest peak in that data is also 16.5 kW, so if there are any spikes higher than that, we know they last less than a half a minute.

What's amusing to me is that every person I've talked to about quoting a generator wants to look at the breaker panel and breaker sizes, which really are not any indication of actual demand of the house. They talk about "shedding this 50A load", just because the breaker to a sub-panel happens to be 50A. Nevermind the fact that I've never even seen 30A loading on that particular sub feed. If you add up all of the breakers in my 7 breaker panels, you'd get a very big number, many times the size of my main breaker.

 

[tradosaurus]

Good advice, as always, Peter! But do note that my 35 kW number is steady state, without considering surge/inrush of each appliance, which could easily be 100 kW for a brief few seconds immediately following a long outage without load management.

Simply put, I summed the steady-state draw of each appliance, in calculating that load. Add maybe 4 kW for lighting, computers, TV's, refrigerators... we're likely at 35 kW max steady state, if all HVAC's, pool heater, well pump and shop air compressor run at the same time. Heck, now that I think of it, I didn't even include kitchen range (10 kW?) or microwave ovens (2x 1.5 kW?). Now we're at 45 kW.

The scenario of all AC compressors, refrigerators, pool heater, and air compressor running simultaneously for minutes or an hour is very likely, immediately following any extended outage. Although with our goal being the avoidance of extended outages, I guess all of those large loads may never run concurrently under normal circumstances.

Maybe see what you absolutely need in an outage. My bet is that you could get by with 20 kW with a plan turning on breakers for the absolute minimum needed for comfort.

 

[vanskills]

I’m a master electrician so obviously did it all myself.

Used a 13000 predator from harbor freight for $1000 dollars

Bought a manual $7 electrical panel switch/lever

Wired it up to a 60 amp breaker in main panel

Probably cost me $1200 all in

Now if you don’t know how to to do that stuff, buy a generac an an electrician to wire it all up…probably cost ya $10,000 all in

 

[WinterDeere]

Maybe see what you absolutely need in an outage. My bet is that you could get by with 20 kW with a plan turning on breakers for the absolute minimum needed for comfort.

That's what I've been doing. The plan is to get away from that, to have a fully-automated system. I won't always be around to babysit the wife and kids, when the power goes out.

Now if you don’t know how to to do that stuff, buy a generac an an electrician to wire it all up…probably cost ya $10,000 all in

Try $25k - $30k, based on initial quotes.

I know how, I'm an electrical engineer and grew up working in building trades, specifically plumbing and HVAC. And if I'd done this in 2023 while my business was still young and not very busy, instead of installing my own in-ground swimming pool, DIY'ing this would have been a slam dunk. But I have no time for it, right now, I'm just too busy with new business. I can make more than the cost of install by spending those same hours at my desk, while also building my business and not having to turn away customers, versus spending those hours to DIY it.

That said, I'm still frugal. Grew up cash-poor, some things you just can't shake. So, I'll be price shopping for best options, as much as I can manage.

 

[ponytug]

@WinterDeere given that time is short, have you considered asking an electrician to bring the right metering / monitoring equipment and cycle everything to get an idea of surge vs steady state? One needs a much better than average DVM to accurately capture peak in rush currents. As you are in the business, you might have the equipment around.

As you have identified that the potential in-rush current might be high due to the number of items that might start simultaneously, can I suggest something? For the less important items, I would put time delay relays set to various delays to allow non-critical, but useful loads to sequentially power up after an outage, e.g. each HVAC, the well pump, air compressor, pool, pool heater, etc. If it were me, I would consider using a small time delay relay set to drive a larger power control relay/contactor. You could also get fancier and use a PLC to control a series of circuits, or get super fancy and use WiFi breakers (not a fan of those personally, $$$, and may not be available for your panels). VFDs on the air compressor/well pump can also have delay starts readily.

Have you considered asking industrial generator installers? You have a business, and the overall size of the undertaking isn't exactly in the normal household size, so the residential operators may be out of their normal comfort zone.

All the best,

Peter

 

[Duds13]

Apologies in advance for a stupid question. You're treating this as a business expense correct?

 

[WinterDeere]

@WinterDeere given that time is short, have you considered asking an electrician to bring the right metering / monitoring equipment and cycle everything to get an idea of surge vs steady state? One needs a much better than average DVM to accurately capture peak in rush currents. As you are in the business, you might have the equipment around.

I have an o'scope, and so could easily buy a fast inductive current clamp to capture inrush currents. Even more easily, I can observe them on my analog Amprobe. But I really don't think it's necessary, we all know motor inrush is going to average ~3x FLA on almost any modern cap start induction motor. Close enough.

For the less important items, I would put time delay relays set to various delays to allow non-critical, but useful loads to sequentially power up after an outage, e.g. each HVAC, the well pump, air compressor, pool, pool heater, etc.

This is exactly what the first installer recommended. I'd been resistant to this idea, because it means pulling wiring out of existing panels and re-routing it to load management modules (LMM's). Space for such hardware, and the wire lengths required to reach them, are both going to be a challenge.

One could say, "this is the installer's problem", but I've seen too many supposed pro's make a mess of such small challenges.

If it were me, I would consider using a small time delay relay set to drive a larger power control relay/contactor. You could also get fancier and use a PLC to control a series of circuits, or get super fancy and use WiFi breakers (not a fan of those personally, $$$, and may not be available for your panels).

No need. There are commercial devices available for this application, with all required functionality already built in.

VFDs on the air compressor/well pump can also have delay starts readily.

Yes, this would be one way to avoid the LMM's, but probably more complicated and more time sunk than the former.

Have you considered asking industrial generator installers? You have a business, and the overall size of the undertaking isn't exactly in the normal household size, so the residential operators may be out of their normal comfort zone.

I have not, because the cost of the equipment they generally deploy is equal to my entire project cost. I was seeing prices of $20k+ for liquid-cooled commercial generators, which is just total overkill for something that might average only 10 - 100 hours of use per year. Whatever I buy, I am quite certain its ultimate demise will be from mice or rust, long before I ever wear it out.

Apologies in advance for a stupid question. You're treating this as a business expense correct?

Since I am operating the business out of my home, I believe I can claim some part of it as a business expense, but I leave that stuff up to my accountant. I am not sure what fraction of the total project cost he will allow me to claim on the business.

Of course, the business is completely self-funded by me, so other than tax implications, it really makes no difference.

 

[not2old]

That said, I'm still frugal. Grew up cash-poor, some things you just can't shake. So, I'll be price shopping for best options, as much as I can manage.

I experienced the same thing prior to turning 40. Before retirement,the thing I considered was paying over 50% income tax on last dollar earned. If the expense wasn't deductible, I would likely make time to do it myself.

 

[ning]

Those prices would pay for a lot of soft start kits and load shedding.
I'd figure on setting up significant load shedding and figuring out which heavy users of electricity get to stay and which get turned off during an outage, and soft start what needs it; my guess is that you can do without at least one or two of those 7 panels, shed their feeds completely, pare things down.
Your propane will last a lot longer.
I'd be really surprised if that house is going to go even 4 days on a 50% full 500 gallon propane tank with everything being available, definitely either manually turn things off when the event happens (harder to tell if the whole house stays running, though without battery you'd probably still notice I suppose) or have the shedding.

 

[WinterDeere]

Those prices would pay for a lot of soft start kits and load shedding.

Agreed. All I can say is that this has been and continues to be a learning experience. I had only the loosest grasp of the economics involved, prior to diving in.

I'd figure on setting up significant load shedding and figuring out which heavy users of electricity get to stay and which get turned off during an outage, and soft start what needs it; my guess is that you can do without at least one or two of those 7 panels, shed their feeds completely, pare things down.

Yes, it appears some load shedding is going to be required for the sake of protecting the generator from potential overload. Given my review of our actual usage versus the size of generator that installers have been recommending (26 - 28 kW), I honestly doubt the load shedding mechansms will ever trip, but it's there to protect the generator from that rare but possible overload.

I learned something interesting today, in that Generac and Kohler reportedly toggle their load management modules differently. The Generac modules reportedly go solely on generator frequency, shedding loads when they notice a drop in frequency indicative of an overload, whereas the Kohlers also monitor current at the generator. The result is that (again... reportedly) the Kohlers will toggle off their first tier of unnecessary loads at 80% generator capacity, hopefully pre-empting any potential overload event. By the time frequency dips long enough for the load management system to determine it's a too-high steady-state load and not just inrush, the argument is that you're likely already overloading the generator.

Again, not a huge issue here, if we're never or almost never hitting those thresholds anyway.

I'd be really surprised if that house is going to go even 4 days on a 50% full 500 gallon propane tank with everything being available, definitely either manually turn things off when the event happens (harder to tell if the whole house stays running, though without battery you'd probably still notice I suppose) or have the shedding.

Every installer tells me to expect 14 days out of an 80% full 500 gallon tank, but since Generac posts fuel consumption data, we can do the quick math on this for a more accurate estimate:

My nice-weather daily-averaged usage rate is 2.5 kW, so starting from a 50% full tank and expecting we can draw down to 15% before any issues, that would give us 9.1 days run time in April or October. More than that, if our tank is above 50% when we lose power, but it doesn't matter, we've never had an outage over 4 days that I can remember.

In July, our usage more than doubles, to 5.5 kW daily-averaged value. That drops our run time to 6.2 days on a 50% full tank, again expecting we can draw down to 15%, maybe a touch less if hot weather affects efficiency. But no worries, our propane company promises delivery in 3 days or less, and I suspect I could always get faster service if really needed.

 

[ning]

We had a "snowpocalypse" here a few years ago ... it was almost laughable how much damage 18" of heavy snow can do, but the trees in the area were very brittle from drought, and many people were snowed in by a combination of the snow and downed trees and power lines; I saw many complaints on ND that people had run out of propane in the first week (most had backup generators) and basically had had to walk out and find somewhere else to stay (and it was barely freezing here after the storm). Some folks were out of propane for well over a month.

There are also load shedding things that will just disable circuits if grid power fails, rather than waiting for the generator to fail at it. My battery failover system (which will someday get a generator connected to it) has connections for tripping load shedders, with the idea that if you're on battery you're not going to be able to run AC. (reminder to self, I do need to set that up, though we rarely lose power in the summer x I don't usually use AC, so the chances of a power failure when I'm running the AC isn't too high).

 

[ponytug]

@WinterDeere, I think that you are underestimating the no load consumption. From experience, propane/gas generators can consume almost the same amount of fuel at 0 (no load) and at 1/4 power loads as they do at 1/2 power. If your particular model has actual data, by all means, use that number. Otherwise, I think that the prudent number to use is the 50% load fuel consumption value. E.g.

Understanding Fuel Consumption In Generators

An essential tool to understanding of fuel efficiency and fuel consumption in generators.

bnhgenerators.com

https://bobistheoilguy.com/forums/threads/fuel-use-vs-load-for-gasoline-powered-generators.370836/

And here, where the natural gas consumption of a 30 or 40kW unit at 1/4 load is about half of the full load.

Fuel Consumption Charts - Generator Source

Check our fuel consumption charts for generators. Optimize fuel use for commercial power needs. Save costs & plan efficiently with our data!

generatorsource.com

My bottom line: use the 50% load fuel consumption number as your low/no load consumption minimum, and anything over 60% of the tank capacity (which is 80% max fuel level) is in the gift category, and not to be relied on.

Just trying to help make sure that the rest of the family is fine if you aren't on deck to tend to things.

All the best,

Peter

 

[WinterDeere]

...people had run out of propane in the first week...

I've lived within 30 miles of the same place for 51 years, and we have never seen an outage longer than 4 days. I'm no saying it couldn't happen, but for any event to have so much impact in a region so used to dealing with everything from hurricanes to blizzards, I'm guessing a generator might not be my biggest problem. Heck, the eye of Hurricane Sandy passed straight over my house, the place looked like a complete war zone, and we still had power restored in something like 2-3 days.

Besides, what's the alternaive solution? Install 2000 gallons of propane capacity, on the off chance we might have the longest-ever-recorded outage sometime between now and when I die? Doesn't feel like a very good gamble, compared to just making reasonable effort to ensure my current 500 gallon tank never dips too low.

There are also load shedding things that will just disable circuits if grid power fails, rather than waiting for the generator to fail at it.

Yes, this is the old way of doing it, and what I had in mind when people talked about shedding loads. But I don't think I would want that here, there simply is no major load I'd be willing to do without for the entire duration of an outage.

If propane consumption and delivery really become a problem, I can always manually shut off some loads, such as my heated and cooled shop or the swimming pool. But avoiding having to regularly shut down and then recover a pool left to stagnate and turn green, or deal with damage to equipment and chemicals in the shop, are a few of the primary reasons I want an automatic standby generator in the first place.

... not going to be able to run AC.

Not run AC in July?!? My wife has rented hotels during outages, for lesser reasons! (I'm not kidding.)

@WinterDeere, I think that you are underestimating the no load consumption. From experience, propane/gas generators can consume almost the same amount of fuel at 0 (no load) and at 1/4 power loads as they do at 1/2 power.

I remember you or someone else saying this previously, so I asked two installers about it this week, and they both disagreed. One guy actually named the no-load consumption at 0.5 gallons per hour on their 26 kW air cooled, which almost perfectly agrees with the 0.49 gal/hour no-load intercept Generac's data yields for their 28 kW Next Gen model.

A simple linear fit on the Generac data gives usage at 0.49 gal/hour + 0.124 gal/hour/kW. But even if we take the absolute worst case of the thing burning it's listed "1/2 load" consumption of 2.23 gal/hour, that's still 78 hours = 3.3 days for your proposed 35% drawdown on the tank from 50% to 15%. And I can't imagine any generator is actually going to burn it's rate listed for 13 - 14 kW load, when seeing an average load of only 2.5 - 5.5 kW... can you?

And here, where the natural gas consumption of a 30 or 40kW unit at 1/4 load is about half of the full load.

Yes, that's pretty close to correct, according to the Generac data. But I think you're drawing the wrong conclusion from that, because:

1. You're talking about natural gas, but the drop-off rate in LP is greater than Nat.Gas. Generac lists their half-load consumption in Nat.Gas at 61% of their full-load consumption, but only 56% in LP.

2. Linear extrapolation of Generac's own data yields a quarter load consumption of 42% of full load for Nat.Gas, but only 34% in LP.

3. My average load is not anywhere near quarter load or half load of the proposed models. It is only 10% most of the year, peaking at 20% in July and August.

Just trying to help make sure that the rest of the family is fine if you aren't on deck to tend to things.

As much as I'm arguing, I appreciate that. I think we'll need to do some test runs to shake things out, and see where we actually stand. I do know that I'll be making an effort to check tank level ahead of any planned travel, and to just put an order in anytime I find it nearing or under 50% full.

I seem to remember my supplier having an up-charge for any delivery under 100 gallons, which is really only 20% of tank capacity. Believing they usually target filling to 80%, that allows me to call them for a fill any time the tank is under 60%, without having to hear from them about any "shorted delivery" hassle. But when push comes to shove, I also don't mind just paying whatever surcharge they'd apply, if that's what it takes to make sure we stay up and running.

 

[groundcover]

Some propane companies are installing wireless monitoring on tanks now, keeps them from showing up for 20 gallons, keeps you from running out.

 

[tradosaurus]

WinterDeere, sounds like you have the money for the purchase and install.
Let's do it and keep us updated!

 

[Duds13]

WinterDeere, sounds like you have the money for the purchase and install.
Let's do it and keep us updated!

I agree! Pull the trigger! Buy once, cry once

 

[WinterDeere]

I agree! Pull the trigger! Buy once, cry once

Yep... already in the works. Had another contractor out last night, just waiting on his bid, now. Once I have that, I'll probably be pulling the trigger.