CALC: Off Grid PV Sizing Tool

Started by MountainDon, January 08, 2010, 09:16:00 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

MountainDon

For quite some time whenever the question of "How big a PV power system do I need?" has come along we've made references to various online calculators, along with the admonishment to begin with adding up how much power you will be using. The key to sizing the system remains, "How much power will you be using."

When I was planning our cabin I wrote a spreadsheet. It worked well but was a little haphazard without much in the way of descriptions to clarify its use. I recently polished it up to make it more publicly presentable. We are making it available here for download with the hope that it proves to be useful to others. The spreadsheet runs in Excel and Open Office. It may not work properly in some versions of Microsoft Works or other less powerful spreadsheet programs.

Download and save the file. The cells you use to enter information are blue. The other cells can not be changed in order to protect the hidden formulas. You can save the file with your own entered data though.

Begin at the top of the form, read instructions.

Part 1 is the energy use inventory. Part A is for 120 VAC use and Part B is for DC use with the voltage user selectable. Part A and Part B both have lines where you can enter a short item description, the quantity, the wattage and the hours of use. This section is set up to have the times entered for the number of hours per week. I find this better as some things may only be run once or twice a week. After the data is all entered the program totals the AC and DC loads, makes an adjustment for inverter loss and then provides a daily power use average.

Appliances will have their power consumption listed on the back or bottom. Normally the power will be listed in watts. Sometimes, like our TV, the label states the voltage, 120 VAC and amps, 1.55 maximum. Multiply the volts by the amps and you have the watts. A more accurate method of determining usage is to use a meter like the Kill-A-Watt, as I've found many items that use less than what the label states.

As data is entered the program will calculate fields automatically. Various fields are copied forward to other sections. There is some sample data pre-entered. This is just for example purposes. You will have to select batteries and PV panels in order to complete the tool.

You are asked to provide the system DC voltage; usually either 12 or 24 volts. This value is the battery bank voltage. That determines the inverter input voltage and the charge controller output voltage. Keep in mind that PV panel output voltages must be high enough to charge the batteries. Panels for use with a 12 VDC battery must have an output at maximum power (Vmp) of about 18 volts or higher. Panels for use with a 24 VDC battery must have an output Vmp of about 36 volts or higher. Of course the charge controller must be able to handle the highest voltage and the highest amperage. That is a topic for a different discussion.


Part 2 is for battery sizing. You will be asked to enter the number of days of autonomy desired, the number of cloudy days reserve you wish to have. You will also be asked to choose the limit of the depth of discharge for your batteries. Remember that a deeper discharge, in other words using more of the battery capacity, will reduce the battery life. Ideally limiting the discharge to 25% is a goal. Many use 50% as their depth of discharge as a compromise between battery life and initial expenditure. You can adjust this figure and the program will adjust the results for the number of batteries required.

The battery section also has an input for temperature correction.

You will also be asked to enter some battery data; battery voltage and amp-hours capacity. There is sample data for the relatively common Trojan L-16RE battery entered as a starting point. The program will provide the number of batteries required for both series and parallel connection.


Part 3 is for the PV panel array. You will be required to enter the average number of sun hours for your location. This would usually be selected for the month of December as that is the most solar deficient month. For a location that is not used much in the winter you may be able to make adjustments that could save on the solar panel expense.

You can find maps for the best estimate of the average number of good sun hours at several websites. I like the maps at...

http://www.nrel.gov/gis/solar.html

Or, the following link which is the older version of the same website...

http://rredc.nrel.gov/solar/old_data/nsrdb/redbook/atlas/

There is a sample map below. At high noon on a clear day, each square meter receives 1000 watts of sun power. If you look at the large yellow areas, you will see that it gets around 6,000 watts on an average day. So, even though the average day is exactly 12 hours, the power you actually get on your panels is equal to about 5 to 6 hours of full sun per day. Since the typical modern solar panel is about 12% efficient, you will get about 700 watts per square meter of panel. So, if the map says that you live in a "six" area, you can expect sun power equal to 6 hours per day over the entire year.

     

Once the sun hours value is entered a selection must be made for the PV panels. There is sample data for a Kyocera 130 watt panel pre-entered. It is suitable for 12 volt systems. There are many others to choose from. Panels must be selected to be suitable for the battery voltage. The maximum power current (Imp) for the candidate PV panel needs to be entered. The program will then calculate the number of panels needed.


It's not a perfect tool but I believe it will be useful and may help someone understand how their potential system will function. You will still have to research batteries and PV panels and other components, but with the calculated data that task may be easier. And of course, you can still ask questions here to help you fill in the blanks.

Criticism, suggestions for improvements and all that are more than welcome. Please feel free to add your thoughts to this topic.

The download link will be in the next message pane.

Just because something has been done and has not failed, doesn't mean it is good design.

MountainDon

#1


EDIT, July 14/2014...

A newer modified version is available. It is attached below this message for download...
PV System Calculator V3.0.xls


Just because something has been done and has not failed, doesn't mean it is good design.


builder_brian

Very nice...thanks for taking the time.... ;D

Whitlock

Make Peace With Your Past So It Won't Screw Up The Present

MountainDon

Dang! I don't know how I did it, but I have an error in that spreadsheet download.  d*

The cell/field for the depth of discharge limit is locked and should not be. When John uploads a replacement I'll post a link to the new corrected file. Sorry.

Just because something has been done and has not failed, doesn't mean it is good design.


MountainDon

#5
Alright now!  :D  There's a new file and it works!

If you downloaded the earlier file please discard it and download the newest one at the above link.

Click here to go there








Just because something has been done and has not failed, doesn't mean it is good design.

OlJarhead

Thanks for all the hard work on this Don!  It's helped me to 'see' things in such a way that I can now begin to plan more realistically.

This is the sort of thing I needed!  I can now plug in something -- say a drill -- and see what it will do to the needed system.

One question I have is this:

If building a system to start out with minimal use in mind (say just weekends with longer stays using the generator to help charge the batteries) would it be prudent to put in a large controller and inverter despite the smaller number of PV panels and batteries if the intent it so to add more later.

For example, with weekend use, very small amount of items and a generator for big ticket stuff (like your set up now more or less) but if the intent is to add a fridge, freezer, dehydrator, vacuum sealer etc etc etc later on down the road and crank usage up would that work?

Essentially I want to be able to add batteries and panels in the future but would like to be able to not have to change out controllers and inverters if possible.

I'm thinking this way becuase I understand that batteries can be spendy as well as panels so say starting with 8 batteries and two panels for minimal use (just an example) but planning to get to 24 batteries and 8 panels in the future with a windmill thrown in is there a controller that would handle that wide range of needs as well as an inverter?

I suspect that the inverter would be good to a point -- like a 2000 watt inverter might be great until you decided to run a miter saw and then you'd need a 4500 watt inverter so in cases like that it won't make sense to try to meet the needs (no catch all inverter right?) but certainly I'm thinking I should be able to install an inverter and controller that can allow the system to grow to a point right?

MountainDon

#7
I'm happy to hear that you have found it useful. That's what I was hoping for.

As to your question(s)...

Choice of charge controller and inverter to allow for system growth.

First, let's say you start with a PV system. Later if you add wind you need a wind power specific charge controller. You can not add the output from the wind generator to the input on the PV charge controller. Wind charge controllers generally have to have a diversion load for times when there is too much power being generated.


Second, the PV charge controller. Having extra unused capacity could be handy when/if the time comes to increase PV capacity. We have extra capacity that would allow adding another series set of panels in parallel with the others, as long as we can fairly well match the existing panels.

When connecting module strings in parallel, you want to match STC (standard test conditions) Vmp and Voc specs reasonably well. The combined array will operate at the whatever string generates the lowest voltage under conditions of use.

For wiring modules in series, you want to match the STC Imp and Isc specs, again reasonably well. Each string will operate at the lowest current specs in the string.


However, another way to add PV capacity would be to add more panels along with another charge controller. That can provide redundancy in the event of a charge controller failure. If you add a second string with a second controller all that matching does not matter.

Some equipment like Outback charge controllers will work together to match the point that each goes into float charge mode. I do not know if that is a big deal or just one of those things are make a nice tidy system.


Thirdly, Inverters. Some (upscale) inverters can be stacked. My Outback could be stacked with another of the same model to either double the combined inverter output watts, or to double the output voltage to provide a 120/240 service. For big systems that can go to four or more inverters.

So it depends on what you are looking at spending initially; cheaper non stackable or more expensive stackable.


Fourthly, Batteries. The experts advise against adding more batteries to a system that is a year or more older. It is said that the new ones will only be as good as the older ones. I think I see the logic. In the real world though many have added new batteries to older ones. Maybe their performance suffers, but it may be hard to quantify.

Ideally one will not have a lot of parallel connections. Three parallel connections is pushing the limit for maximum battery life and performance, according to the experts. I think I've said someplace before that ideally one will assemble a battery system using the fewest number of cells. That's ideal, but there are many systems that are not ideal.


Fifthly, generator. Our battery bank and inverter have enough capacity that we have not had to run the generator for auxiliary power, unless I was to get a big project going and needed to use a large AC tool for extended time. I've used it for equalizing, but even there the PV panels had enough capacity to perform that job back in late summer. Our inverter can actually supply more power than the generator, for short periods. The generator is limited to about 20 amps while the inverter can supply 30 amps.
Just because something has been done and has not failed, doesn't mean it is good design.

OlJarhead

Thanks again Don...with luck I'll be getting the first components of our solar system within the next 6-8 weeks :)

Really depends on what we decide to do with the funds available but that might be one of the routes we take (original plans had it maybe 2-3 months out).

All great info here and I'll have to do some serious thinking and using the spreadsheet makes that much easier :D


davidj

Thanks for this tool - it's great.  At work I deal with vendors trying to sell me computer equipment and sometime they have sizing tools and forms.  Yours is better than anyone fighting for a $20K sale has come up with!

I've put a copy of my completed sheet up on my website in case anyone wants another example.

The summary is that, for a 20x30, all 120V, with a refrigerator and a 1.5HP well pump plus the usual stuff, 8x175W panels and 8xL16P batteries seems to work (which is good, as my inverter is 48V).  In long storms in Winter it's likely I'll need to hook up a generator for an hour or two a day as the batteries look like the weakest link.


MountainDon

Thanks David. Glad it was of some help.
Just because something has been done and has not failed, doesn't mean it is good design.

dablack

I hate to be this guy but your math is wrong. 

To correct for a 90% efficiency, you should divide by .9 not multiply by 1.1. 

Your spread sheet adds up all the watt-Hours/Week that you need to come out of the inverter.  What you want to know is what needs to go into the inverter to get this calculated output.  We are assuming 90% efficiency.

So we will call the input into the inverter "I".  The output is already summed in your spreadsheet but for the formula we will call it "O"

I times .90 = O

We already know O so to solve for I we need to divide both sides of the equation by .90.

So:

I = O/.90

Your spread sheet should divide the summed watt-hours/week by .90 to calculate what needs to go into the inverter. 

With all this being said, I've never designed a PV system and might not understand your spreadsheet or now an inverter works, but I am assuming you only get out of an inverter 90% of what you put in. 

Sorry.  I'm a mechanical engineer and couldn't let it go.......

Austin

MountainDon

Well goodness, you are right. Thanks for the input. That is an easy trap to fall into; I should have known better. The good thing is that it's a relatively small error and in most cases will not affect the outcome when the round up and round downs are done to calculate the # of batteries.  I'll redo the sheet and make note of that when the new one is uploaded.

Inverter efficiency varies a lot from brand to brand and by degree of loading, so there are many variables in any calculation method used to arrive at system requirememnts.
Just because something has been done and has not failed, doesn't mean it is good design.

dablack

I should have noted in my post that the difference is in the noise and the spreadsheet works fine as is.  I just can't help myself when it comes to math.  It really is a very useful spreadsheet and it helped me understand PV systems just by going through it. 


MountainDon

I'm happy it helped you. I have run off a corrected version and will soon have it available.

You're right in that it is a very small difference, but I should have caught it. It's similar to giving discounts on merchandise. If something is discounted 15% and then you get an extra 5% off it is not the same as getting 20% off the original price.  d*

Just because something has been done and has not failed, doesn't mean it is good design.

MountainDon

Just because something has been done and has not failed, doesn't mean it is good design.

OlJarhead

Question:  If I run the spreadsheet with 34v panels and ~7.5imp the sheet is telling me I'd need MORE 270 watt panels then 18v panels (210watt) at 11.43imp when calculating for a 12 volt system.

I found this interesting as I was trying to help a friend but the spreadsheet kept telling me that I needed a LOT more 270 watt panels then 210's.

Then I changed the bus voltage to 24vdc and it cut the panels way down.  I then realized that the spreadsheet doesn't really account for panel voltage -- or maybe I'm confused??

If I'm running a Xantrex C40 and I'm pushing in 70vdc with two 270 watt panels I 'should' have more charging power then running 4 210 watt panels hitting the Xantrex at 60vdc I'd think -- but since the sheet assumes an 18vdc panel for a 12vdc bus how do you figure what to do?

I'm basically blabbering now aren't I? hahahaha  d* d* d* ??? ??? ???

MountainDon

There is no place to enter PV module voltage.

There is the system voltage which is the battery system voltage; cell G66 if we have the same version.

When it comes to choosing panels one must do some work. Rows 135 - 140 and 164 -166 give some clues.  Basically the modules, or pair or more in series, need to have a Vmp approximately 1/5 times greater than the battery bank voltage. Choosing the charge controller, fuses or breakers and other components is beyond the realm of the tool but remain factors to be considered.

In your example, using a module with a Vmp of 34 volts should not require any panels in series, so in cell F167 you should enter numeral 1.  If you wanted two modules in series for any reason you would enter 2.  And at that point my calculator takes the easy way out and tells you that you need more panels than really necessary (in the example I ran for myself with 1 panel in series it needed 3 in parallel. Then I changed the series number to 2, just because I wanted the higher voltage... and the calc told me I needed 3x6 = 6 modules. Which is not true to get the desired watts. So I have to rework that section some time. If you understood that.... great. If not, think about it and I will get that sorted out and try ti make it clear.     The correct answer would be 4 modules and would supply more watts than needed....    I may need some "if.... thens....." or a verbal set of instructions.

If you want to save and email me the filled in form with the basic choices of panels, batteries in a note or something I'd look at it.



Just because something has been done and has not failed, doesn't mean it is good design.

OlJarhead

Thanks Don,

I left the panels set to 1 in series because the 18vmp and 34vmp panels are both more then enough for a 12vdc system.

However since the calc assumes 18vmp for a 12volt system setting the imp for the 34vmp panels shows something entirely different then what one would expect -- meaning that if you put 7.x imp for the higher voltage panels the calc automatically assumes that from an 18vmp panel and tells you that you need twice as many panels then a lower wattage 18vmp panel running 11.ximp

Make sense?

SO my question was this:  would not the 34vmp 7.ximp 270watt panel have more charging power then the 18vmp 11.ximp 210 watt panel?

And if so how can you change the calcs to show this?



MountainDon

A first look and quick think on that reveals there is an issue.  The way the sheet calculates the answer is correct if one is NOT using an MPPT charge controller. With a non MPPT controller any excess power because of the voltage from the PV modules being higher than needed is wasted. So the spreadsheet is only considering the amps, the Imp of the panel. Therefore the calc comesup with three of the 210 watt (11.23 Imp, 18 Vmp) or 4 of the 270 watt (7.45 Imp, 35 Vmp).    Note my numbers are with the quick estimated use and default L16 battery info, so YMMV.

If an MPPT charge controller is used everything changes. So I need to give this some thought and make some changes for the theoretical use of an MPPT controller (which I dearly love).

Thanks for noticing and letting me know.

BTW, why are you even looking athe the 210 watt modules?  There is a otation on the web page that states: "Note: These solar panels cannot be imported to US, Canada, Mexico, Puerto Rico, nor any of the United States territories."  Probably because they are "non-UL, Tuv or CE listed" and perhsaps hazardous when in use.   ???

Just because something has been done and has not failed, doesn't mean it is good design.

MountainDon

WARNING: An issue has been identified in the last section of the calculator; the PV array. It works fine for modules that are series wired to bring low output panels up to the minimum required for charge controller use.

That is, a 12 volt battery system should have a panel output voltage total of about 17 to 18 volts and a 24 volt battery system should have around 34-36 volts from the PV modules. The calculator fails in sizing the array when the user is planning on using modules with say 34 volt ratings on 12 volt battery systems.

We'll see about a fix soon as we can.



Just because something has been done and has not failed, doesn't mean it is good design.

Squirl

This seems to fit into a topic recently in the off-grid power thread.  With chargers like the C40 that do not have MPPT technology the additional voltage above 1/5 more than the battery bank is wasted.  So if you have a 12v battery bank, your charging voltage would be around 14v and your panel voltage should be round 17-18v max.  Anything over is wasted power (money).  So with a C40 and a 12v bank you would max out at around 480 (12x40) watt panel array. All are rough estimates.

Sorry if I was a little redundant of don, I wanted to answer to your question specifically about the C40 controller.

OlJarhead

#23
Thanks Don, I look forward to your thoughts on this and continued work.

I used the 210's as an example only as I was in a rush.  I have the 205's and wanted to compare an 18vmp panel to a 34vmp one is all.


Squirrel -- I see voltage in another way ;)  Running under 48vdc from the panels to the charge controller limits range on 10awg wire to under 50 feet (not sure the distance) where as 48vdc will run 50 feet on 10AWG according to xantrex.  In my case I'm pushing 60vdc from the panels which gives me a little better (I think) performance at 50 feet over 10awg.

I'm no electrician mind you, I'm just thinking that running 18vdc from the panels would cost me in wire size (read $$$).  

So, question:  and I should post this in the other thread, but here goes anyway:  Am I losing something by putting the panels in series and pushing out 60vdc to the controller?

Another question is this:  If I put 3 batteries in series I don't change the AH rating just the voltage so if I put 3 panels in series I'm wondering if all I'm doing is increasing voltage and not wattage?  I figured 3 205watt panels would give me 615watts of charging power.

Anyway, good topic!


EDIT: I copied the part above that I italicized and pasted it into the OFF GRID thoughts....  topic as it fits better there. It might be nice to try to separate the thoughts.


OlJarhead

Here's a question:  when using the tool and putting in the amount of daily sun (according to the chart) what exactly does that mean?  After all, with my MPPT controller I'm seeing over 10 hrs of charging power per day (usually 12-13hrs).