Maker Pro
Maker Pro

More Batteries vs Running Generator

B

Brent Geery

Jan 1, 1970
0
I've been 100% solar for about 8 years now, without any backup
capability, and do just fine. However, when I build my home, I'll
have a much larger system to deal with, so economics are going to have
to get much closer scrutiny, and I will have to finally add some sort
of backup generation capability.

Now my question is, how do you determine the best economic balance
between adding additional battery capacity vs the cost of running and
maintaining a generator? I'm not aware of anyone that actually tries
to calculate the most economic balance.

I did some calculations, and it's no surprise, that adding more
reserve capacity to the battery bank, increases the average lifetime
per kWh cost.

For example, assume a daily average demand of 10 kWh, and you have a
20 kWh battery bank that cost $1,200, and it provides 1.5 days
capacity, and will last 7 years with a normal daily DoD of 50%. This
energy will cost an average of $0.047 per kWh.

If we double the battery bank, to give 3 days capacity, we will
increase the bank life to about 10 years, lower the average DoD to
25%, and raise the cost battery bank cost to $2,400. Now, this energy
will cost an average of $0.066 per kWh. That's an increase of just a
hair under 50%!

If we double the battery bank again, to give 6 days capacity, we will
increase the bank life to about 15 years, lower the average DoD to
12.5%, and raise the cost battery bank cost to $4,800. Now, this
energy will cost an average of $0.088 per kWh. That's almost double
the per kWh costs for one day!

Now to figure the cost per kWh for generator created energy. There is
the cost of the generator itself, then maintenance costs, then fuel
costs. Let's assume we buy the Honda EG3500 for $1000, and that it
will last 10,000 hours, and that maintenance costs are 50% of the
purchase price, over the life of the generator. It's rated at 3000
Watts 125 VAC at full load, and we will assume the AC-to-DC conversion
of the Trace inverters battery charger function is 90% efficient. The
generator uses 0.5 gallons of gas per hour, and we will assume a gas
cost of $2.25 per gallon. Finally, we will assume that when we have
to charge with the generator, we will only give the batteries a bulk
charge (batteries up to 80% charged,) to avoid the long and
inefficient acceptance charge run times. That gives us a grand total
of $0.567 per kWh over the life of the generator.

So, how do we determine the point it's cheaper to run the generator
than increase the battery bank capacity further? To complete the
analysis, I guess, we would need to determine/estimate (some how!) the
number of times a year that there is multiple extended cloudy days.
But how to get this information?

For example, I might discover that there at average 40 times a year,
that there is extensive clouds extending more than one day, ten times
a year with extensive clouds extending more than two days, and 3 times
with extensive clouds extending more than 4 days.

So, using the above weather examples:

40 times yearly of >1 days clouds, if I have a 1.5 day battery bank
capacity, would require 400 kWh a year of generation at a cost of
$226.80, or $1,587.60 over the 7 year life of the 1.5 day capacity
battery bank.

10 times yearly of >2 days clouds, if I have a 1.5 day battery bank
capacity, would require 200 kWh a year of generation at a cost of
$113.40, or $793.80 over the 7 year life of the 1.5 day capacity
battery bank.

4 times yearly of >3 days clouds, if I have a 1.5 day battery bank
capacity, would require 120 kWh a year of generation at a cost of
$68.04, or $476.28 over the 7 year life of the 1.5 day capacity
battery bank.

Adding these up, in our example, give a total generator operation cost
of $2,857.68 over the 7 year life of the battery bank. This is more
expensive (by $457.68) than the cost of doubling the battery bank to 3
days of capacity, so would not be the best economic choice in this
example.

Using the same example, but using a battery bank with 3 days capacity:

40 times yearly of >1 days clouds, $0, as the battery bank carries us
through it.

10 times yearly of >2 days clouds, $0, as the battery bank carries us
through it.

4 times yearly of >3 days clouds, if I have a 3 day battery bank
capacity, would require 120 kWh a year of generation at a cost of
$68.04, or $680.40 over the 10 year life of the 3 day capacity battery
bank.

$476.28 is cheaper than the $1200 it would take to give us another 15.
days of capacity (4.5 total then,) thus would be the most economical
choice in this example.

Comments please. I my next message, I'd like to debate adding PV
capacity vs running a generator.
 
W

William P.N. Smith

Jan 1, 1970
0
Brent Geery said:
For example, assume a daily average demand of 10 kWh, and you have a
20 kWh battery bank that cost $1,200, and it provides 1.5 days
capacity, and will last 7 years with a normal daily DoD of 50%. This
energy will cost an average of $0.047 per kWh.

If we double the battery bank, to give 3 days capacity, we will
increase the bank life to about 10 years, lower the average DoD to
25%, and raise the cost battery bank cost to $2,400. Now, this energy
will cost an average of $0.066 per kWh. That's an increase of just a
hair under 50%!

What if you adjust the DOD to keep the battery lifetime constant, how
does that effect your numbers? What if you reduce your battery
lifetime?
costs. Let's assume we buy the Honda EG3500 for $1000, and that it
will last 10,000 hours, and that maintenance costs are 50% of the

Yikes! I suspect you are off by a factor of at least 10 in your
generator lifetime. Dunno what the Honda specs are, but that style of
generator generally has a lifetime measured in hundreds of hours, not
tens of thousands.

That said, generator power is probably going to be a _lot_ less
expensive than adding more solar power or bigger batteries, but who
wants to run a generator 24x7?
of $0.567 per kWh over the life of the generator.

Note that you are comparing storage with generation, which is kinda
apples and oranges.
 
Brent Geery said:
...assume a daily average demand of 10 kWh, and you have a 20 kWh
battery bank that cost $1,200, and it provides 1.5 days capacity,
and will last 7 years with a normal daily DoD of 50%. This energy
will cost an average of $0.047 per kWh.

Sounds cheap.

Trojan rates batteries in "Lifetime Energy Units," ie the total amount of
energy that a battery can store over its lifetime. This number increases
with DOD. The T-105 DOD vs #cyc curve shows about 750 cycles at 100% DOD
(LEU = 473 kWh), 1000 at 70% (441), 1500 at 42% (397), 3000 at 20% (378),
and 5500 at 10% (347.)

At 50% DOD, a T-105 might store 397+(441-397)(0.5-0.42)/(0.7-0.42) = 410 kWh,
If the battery costs say, $60, ignoring shipping cost, cables, interest on
the investment, shelf-aging, the cost of conditioned space, inverter loss,
and so on, that's $60/410 = 0.146/kWh.
If we double the battery bank, to give 3 days capacity, we will
increase the bank life to about 10 years, lower the average DoD to
25%, and raise the cost battery bank cost to $2,400.

At 25% DOD, a T-105 might store 378+(397-378)(0.25-0.2)/(0.42-0.2) = 382 kWh,
at $60/382 = 0.157/kWh. If the larger bank is charged with a larger generator
that also powers the load when it is running, it might end up storing more
electricity over its lifetime. (Imagine the limit, with no batteries, and
a generator that runs all the time.)
...To complete the analysis, I guess, we would need to determine/estimate
(some how!) the number of times a year that there is multiple extended
cloudy days. But how to get this information?

A TMY2 simulation.

Nick
 
M

m Ransley

Jan 1, 1970
0
A Honda EG 3500 , 3600 rpm, 10000 hrs? I think you would be lucky to
get 3000 hrs more like 1500-2000
 
B

Brent Geery

Jan 1, 1970
0
A Honda EG 3500 , 3600 rpm, 10000 hrs? I think you would be lucky to
get 3000 hrs more like 1500-2000

Really? Home Power was using a alternator connected to a 5 HP Honda
Industrial/Commercial series engine, and they said that it had
something like 5,000 on it, and still going strong, according to one
of the old articles from the Solar 1 CD. Note, that Honda's
Industrial/Commercial engines are nothing like those crappy Tech and
Briggs models!

But you are right, that does seem like a damn long time! 15,000 hours
would be about 900,000 miles at 65 MPH! Maybe because HP was only
running their engine a little over idle, due to the relatively small
load of a 100 amp alternator?

Do you have a suggestion for the overall cheapest generator to
own/operate on a per kWh basis? I'll use that instead.
 
B

Brent Geery

Jan 1, 1970
0
What if you adjust the DOD to keep the battery lifetime constant, how
does that effect your numbers? What if you reduce your battery
lifetime?

How would I do that? I'd have to adjust my average daily load.
That's a fixed number (well, it can vary be season, but you know what
I mean.)
Yikes! I suspect you are off by a factor of at least 10 in your
generator lifetime. Dunno what the Honda specs are, but that style of
generator generally has a lifetime measured in hundreds of hours, not
tens of thousands.

Honda's industrial/Commercial engines are not the same crap you find
from Briggs or Tech. Home Power's DC generator used a Honda I/C
engine with 5,000 hours on it, and was still going strong, with no
major work needed. Were they just lucky? In any case, fuel costs
dwarf the costs of the generator itself! :(
That said, generator power is probably going to be a _lot_ less
expensive than adding more solar power or bigger batteries, but who
wants to run a generator 24x7?

Just in gas costs alone, small generator created energy it horribly
expensive, and that's assuming loading the poor thing to it's most
efficient max. There is a reason why people choose PV, and most of it
is about economics (or lack thereof) of running a generator as the
primary generation source. :)
 
B

Brent Geery

Jan 1, 1970
0
Brent,

www.nrel.gov/homer

Homer allows you to vary many of these parameters and come up with
reasonable economic scenarios. Download Homer Pro -- it's free although
registration is required.

It can be a bit tricky to set up your generator/charger parameters. And I
don't believe it accounts for the various ways in which a charger can be
programmed. But I've found it quite useful in looking at the economic
tradeoffs you're considering.

It does seem to underestimate generator run time, at least the way I have
it set up. I suspect this is due to not taking into account the fact that
with a constant voltage charging scheme, current into the batteries falls
off as the SOC increases.

In my case, it turned out that adding extra electrical generating capacity
(PV) to my wind system became economically feasible when propane was at a
certain price.

In my case, I'm adding more PV than the program suggests, as we intend to
add some items that will increase our electrical consumption, and I'd also
really like to minimize generator run time during the summer (when the wind
is weakest here).

Thanks for the pointer, Ron! Very cool program, but I already see a
few things I don't like about it. For example, I can't figure out the
grid power settings. But, I'll play with it.
 
W

William P.N. Smith

Jan 1, 1970
0
Brent Geery said:
How would I do that?

By putting in a battery bank only (say) 50% larger than you have now,
increasing it's DOD, and still getting three days of capacity.
Just in gas costs alone, small generator created energy it horribly
expensive

Yeah, my mistake, I was thinking of diesel generators, which can cost
a lot less per KWHR, but have much longer lifetimes, and larger
upfront costs...
 
B

Brent Geery

Jan 1, 1970
0
What is it about the grid power settings that you don't understand? Since
I'm off grid, I haven't played with the grid-connected options.

If I can't help, the authors have been responsive in the past.

OK, things I don't like about HOMER:

PV settings-

Mounting and charge controller costs are clumped together. These
things don't scale linearly with system size. Adding one extra panel,
may require adding a completely new mount or controller, that could
handle many more panels otherwise, raising the $/KWh cost of the
system. Also, there is no easy way to directly compare between
tracking and non-tracking mounts.

Grid Settings-

And my biggest problem with Homer...
There is no way to adjust these options to account for Time-of-Use
metering, combined with net metering (only net metering.)

Also, I don't see a way to account for the way California has its net
metering set up, with sell-back limited to the net yearly charges you
paid for that year. Meaning, the best you can ever do, is have a $0
electric utility bill for any year.

I'd really like to experiment with grid-tied Time-of-Use net metering,
especially as an alternative to an expensive battery bank, but I can't
with the current version of HOMER.
 
B

Brent Geery

Jan 1, 1970
0
The above paragraph, no longer applies, as I did figure out how to
make California style net metering work, by setting the "sell-back"
setting to "0". However, time-of-use is still not able to be modeled.
Those are good points. As neither of those considerations apply to my
setup, or projected setup, I've not thought about them.

The reason I'm interested in net-metering, is to determine if
extending the grid would be worth it or not.
Perhaps if you emailed Dr. Lilienthal, he would consider incorporating
those features.

I think I will, thanks, Ron.
 
M

m Ransley

Jan 1, 1970
0
generator life depends alot on the load you put on it, a light 10 %
load wont wear it out nearly as fast as a 90% load. My 7500 13500 w
generac uses apx 1 gal hr at full load and at 10 % will run 5 hrs on 1
gal. Figure running your car at 3600 rpm near full throttle it may
last 50000 miles, at 1800 rpm at 55 mph maybe 250000.miles . Generac
with steel sleeve and pressure lube and oil filter I beleive are as good
as Honda, and the Honda has no oil filter.
Cost depends on how many hours you will need to run it a year . For 4
to 6 thousand a 1800 rpm diesel may last 20000 to 80000 hrs. 10 to
40 times as long as the honda for 4 to 6 times the price.
 
B

Brent Geery

Jan 1, 1970
0
I was just speaking with Dr. Lilienthal about some other issues, and I
brought up the problems you mentioned.



To handle this non-linear data in the model, you can enter separate line
items for the different size PV systems in the Costs spreadsheet area,
rather than entering a single cost and scaling it.

Duh, why didn't I think of that!

Of course this still creates an odd situation when dealing with parts
that have different estimated life times, but that can be somewhat
worked around by manually doing some precalculations. For example, I
have to divide 1 1/3 controllers (15 year life) or 1/2 the cost of the
mounts (40 year life) into the cost of a 20 year PV panel. Would be
easier (and probably a little more accurate, taking interest rates
into effect) if they had their own lines.
I didn't ask him about this, but it looks to me as if you would have to run
different simulations. Consider forwarding this to him as a suggestion.

That's what I'm doing now. Interestingly, In all the test's I've run,
2-axis tracking does not pay. This was a surprise to me, as the solar
data I have, shows a much higher PV production with 2 axis tracking
(almost double!)
This is high on their priority list to be added. No time frame is
available, though.

Great! At least Dr. Lilienthal is aware of the desire for that
feature. It really is a very cool bit of software.

One other question, that you (or Dr. Lilienthal) might be able to
answer:

How can account for the cost of inflation, but not the cost of money?
In other words, assume that I can't invest the money, for whatever
reason (that's my situation), but I still want to take into the
account the increases in costs caused by inflation.

Do I just enter "0" for the interest rate? The inflation rate? Or
even a negative number? Right now, I'm entering the inflation rate,
but I'm unsure if this is really giving the effect I desire.
 
W

William P.N. Smith

Jan 1, 1970
0
generator life depends alot on the load you put on it,

And the initial quality of the engine. Anyone know how the B&S
Vanguard engines rate, or what their service life might be? Except
for the usual maintenance, the only thing in the service manual is
"Clean combustion chamber deposits" every 500 hours (any idea how that
might change with synthetic oil and natural gas fuel?).

The one in my generator's got 76 hours on it, so I'm not too concerned
yet...

Thanks!
 
M

m Ransley

Jan 1, 1970
0
You wont have deposits with NG. and oil will run cleaner. Lower ends
last longer but gas is dry and is harder on the upper end, piston rings
cilinder and valve guides etc. Does it have a steel cilinder and
pressure oil delivery and oil filter?
 
B

Bughunter

Jan 1, 1970
0
And the initial quality of the engine. Anyone know how the B&S
Vanguard engines rate, or what their service life might be? Except
for the usual maintenance, the only thing in the service manual is
"Clean combustion chamber deposits" every 500 hours (any idea how that
might change with synthetic oil and natural gas fuel?).

The one in my generator's got 76 hours on it, so I'm not too concerned
yet...

Thanks!

I have a 5k watt generator that uses the B&S Vanguard 9Hp. It runs on
propane.
I have had this for 9 years. I have to admit that I was not very dilligent
in changing the oil,
but it never seemed to get dirty.

The generator did not get what I would call high usage, maybe 10-15 hours a
week for the summer months.
Two years ago, I ran it 5-8 hours a day, 6 days a week while I was building
a house for about 4 months.
Most of the time, it was running very heavy loads including a 4hp compresser
and other power tools
while also intermittantly running a 40amp battery charger. I was pushing
that little guy to the max.
It does not have a hobbs meter, so I have now way of knowing the actual
hours on it are.

At the end of that building season, it started running rough, and finally
would not start at all.

I brought it home, thinking I'd have to put a new engine on it. Whle working
on it, I found a tag on it indicating
that it was 9 years old. Geeze, I could remember changing the oil in in only
a couple times. A consequence not
making a convinent drain arrangement after putting it in an enclosure house.

I took apart the head and found that the valves were encrusted with carbon
to a point where the you could
not crank it. After scraping off the carbon and polshing them up a bit, it
ran again like a charm. That was the only maintenance it ever needed.

I expect I'll be able to get another 9 years of use and abuse out of it.

I was impressed with how easy it was to pull off the valve covers, and by
the fact that they used
a stainless steel (or similar material) for the gasket. I didn't even have
to replace any gaskets!

The higher HP Vanguards have a pressureized oil system with a filter. I
expect that would be even
more reliable than the 9hp (which does not).

The Vanguard line is made by Mitibishi for B&S. It's a damn good little
engine, simple, reliable and capable
of surviving years of abuse.

I don't know what the ratings are, but I'm impressed with my sample of one.
 
Top