Reasons why my system is underperforming

lesyl

Greetings,
I have a 4.14kw system made up of 18 Yingli 230W panels and an SMA 4000US Inverter. The switch was flipped on the system on Dec. 6, 2010. The estimated minimum annual output is 7,858 kwh/year. The size of the system and estimated output is within the range of two other estimates I received. As of this morning, three weeks short of the 1 year anniversary, the system has produced 6,388 kwh. Clearly it’s going to fall well short of the estimated minimum.

Late last week, the company that did the installation sent out a technician (one of two who did most of the installation) at my request to go through the system. He spent about two hours going through it. When he was finished, he said he had all the numbers, everything was producing, and that the owner of the company would call this week to discuss the numbers.

I did a lot of asking and looking around before I decided which companies to deal with. The company that installed the system has an excellent reputation here in Tucson.

I realize there could be a perfectly valid explanation, but I’ve lived long to know a healthy dose of skepticism is often a good thing. In case the owner tells me everything is fine, I’d like know what questions I should make sure to get answered. I don’t necessarily know enough about the technical aspects of how solar works to know what to ask and that the answers are on the up and up.

So, any thoughts? What possible reasons could cause a system to under perform? Obviously malfunctioning equipment or a mistake on the installation is on top of the list. Is there anything else? What questions should I make sure I get answered?

Thanks, Tim

ggunn

Re: Reasons why my system is underperforming

lesyl wrote: »

Greetings,
I have a 4.14kw system made up of 18 Yingli 230W panels and an SMA 4000US Inverter. The switch was flipped on the system on Dec. 6, 2010. The estimated minimum annual output is 7,858 kwh/year. The size of the system and estimated output is within the range of two other estimates I received. As of this morning, three weeks short of the 1 year anniversary, the system has produced 6,388 kwh. Clearly it’s going to fall well short of the estimated minimum.

Late last week, the company that did the installation sent out a technician (one of two who did most of the installation) at my request to go through the system. He spent about two hours going through it. When he was finished, he said he had all the numbers, everything was producing, and that the owner of the company would call this week to discuss the numbers.

I did a lot of asking and looking around before I decided which companies to deal with. The company that installed the system has an excellent reputation here in Tucson.

I realize there could be a perfectly valid explanation, but I’ve lived long to know a healthy dose of skepticism is often a good thing. In case the owner tells me everything is fine, I’d like know what questions I should make sure to get answered. I don’t necessarily know enough about the technical aspects of how solar works to know what to ask and that the answers are on the up and up.

So, any thoughts? What possible reasons could cause a system to under perform? Obviously malfunctioning equipment or a mistake on the installation is on top of the list. Is there anything else? What questions should I make sure I get answered?

Thanks, Tim

When you say "estimated minimum output", how was that determined? If it was with PVWatts, check what derate value was used; sometimes installers use an optimistic number to sell systems. The default is 0.77, and if the derate used was over 0.8 I would ask the installer for a justification for it.

Understand, too, that PVWatts operates with a +/- 10-12% uncertainty and what you are seeing is pretty close to the lower bound of that uncertainty. It could be something as simple as there being more cloudy days than usual over the past year.

mike95490

Re: Reasons why my system is underperforming

How often, outside the rainy season, do you wash your panels down ?

A slight layer of dust film will only drop production a couple watts, but over time it adds up.

lesyl

Re: Reasons why my system is underperforming

Duh...I have no idea how the output estimated was determined. So, that's a question to ask. However, the estimates were all remarkably similar despite each company using different brands of panels and converters.

Our yearly rainfall is a little below average so I doubt we've more cloudy days than normal, but our summer was even longer and hotter than our usual long, hot summers. I know heat hurts output.

Thanks ggun

lesyl

Re: Reasons why my system is underperforming

Mike

I keep an eye on them and wash them when they're dirty. I try to keep them as clean as possible.

drees

Re: Reasons why my system is underperforming

FWIW - PVwatts estimates 6884 kWh for an 4.1 kW south facing system mounted at 32* in Tuscon.

7800 kWh/year might be optimistic?

Reasons for shortfall:

1. Weather (hotter or cloudier than normal - do you have monthly output to compare?
2. System not facing south or not mounted at sub-optimal angle.
3. Shading

Would be helpful to have monthly output and to know how other systems in Tuscon are doing compared to PVwatts this year.

lesyl

Re: Reasons why my system is underperforming

Thanks drees. That's disappointing to see the PVWatts estimate. We thought we were paying for a system that would produced in the neighborhood of 7,800 -- and even that number we were told was conservative.

I screwed up by not tracking monthly production. I know that would help.

Shading is a none issue. The system faces south and is set at the optimal angle. As I mentioned, summer was even hotter and longer than our usually long, hot summers. That's the one thing...

ggunn

Re: Reasons why my system is underperforming

lesyl wrote: »

Thanks drees. That's disappointing to see the PVWatts estimate. We thought we were paying for a system that would produced in the neighborhood of 7,800 -- and even that number we were told was conservative.

I screwed up by not tracking monthly production. I know that would help.

Shading is a none issue. The system faces south and is set at the optimal angle. As I mentioned, summer was even hotter and longer than our usually long, hot summers. That's the one thing...

That's why it pays to do your homework and double check any claims made by folks wanting to sell you stuff. PVWatts is extremely easy to use and can generate a pretty good forecast of yearly output from minimal system details. All you need is your location, the DC rated system size, the tilt and the azimuth, and it spits out a pretty good number in a few seconds.

Tracking monthly production after the system is in place would have only helped by lowering your expectations.

mlail

Re: Reasons why my system is underperforming

As someone pointed out, dust is a problem over the long run. I know the Pheonix area was hit with dust storms. As I understand it, your area is not as prone this this but without a regular rain to clean off the panels, you might want to clean them manually.

solar_dave

Re: Reasons why my system is underperforming

mlail wrote: »

As someone pointed out, dust is a problem over the long run. I know the Pheonix area was hit with dust storms. As I understand it, your area is not as prone this this but without a regular rain to clean off the panels, you might want to clean them manually.

The garden hose was my friend many times this summer, out at dawn with a ladder to reach them all in 2 setups. It helps the production as much as 5% IMHO.

a0128958

Re: Reasons why my system is underperforming

It looks like the heat in Tuscon significantly affects the output.

Using the same PVWatts data noted above, average daily solar hours in Tuscon is 6.59 (25% more than what we have here in Dallas, BTW).

Multiply this times 365 days/yr and multiply times 4.14 kW, you get an estimated 9958 kWh/year. This is the 'nameplate' estimate for a 4.14 kW DC system in Tuscon. It takes into account all factors that can influence insolation (clouds, humidity, haze, time of year, distance to the sun, dust on panels, etc) for Tuscon, but, does not take into account temperature (in fact it assumes STC temp).

The estimated 9958 kWh/year assumes default PV system design (array pointed due South, tilt fixed at latitude, no shade) and default equipment/wiring performance. This is the commonly refered to PVWatts 'default' 0.77 derate factor.

Looking again at PVWatts, it shows an estimated 6651 kWh/year AC Energy when taking into account temperature. 6651 kWh/yr is one-third (33%) lower than the nameplate estimate. Looks like Tuscon is a dry and generally cloudless climate (high nameplate estimate), and it's hot (much lower AC Energy estimate).

Again, this 6651 kWh/yr estimate continues with the 0.77 default derate factor.

Your actual of 6388 kWh produced on 11/14/10 is 22 days short of 365 days of production. Prorating your 6388 kWh to a full year yields 6798 kWh/yr.

6798 kWh/yr actual is 2% better than the Tuscon default estimate of 6651 kWh/yr. So you slightly beat a conservatively derived estimate for your first year.

Going one step farther, if the PVWatts AC Energy estimate is 6651 kWh/yr at a 0.77 derate factor, then if you divide 6651 kWh/yr by 0.77, you get 8637 kWh/yr if you assume the equipment/wiring operates at theoretical perfection in addition to the already assumed perfectly facing South array that's tilted at latitude that doesn't have any shade. Taking your actual of 6798 kWh/yr and dividing it by 8637 kWh/yr, you get an estimated derating factor of 0.79. Or said another way, your system is 79% efficient across an entire year, 2 points better than default.

The estimate provided to you, 7858 kWh/yr, is 18% higher than the Tuscon default estimate of 6651 kWh/yr.

Since you had three contract proposals all with the same 7858 kWh/yr estimate, that would lead one to conclude that some tool other than PVWatts version 1 was used to arrive at the provided estimate, and that this tool was used for all 3 proposals. For example, maybe PVWatts Version 2 was used to do an estimate to your exact location, and your specific location has some characteristic different from the Tuscon location used in PVWatts' Version 1.

If we assume the 7858 kWh/yr contract proposal estimate is indeed correct for your location, and assuming it was done based on a 0.77 derate factor, then if you divide this by 0.77 you get 10205 kWh/yr, again assuming the equipment/wiring operates at theoretical perfection in addition to the array facing due South, tilted at latitude, and without shade. Taking your actual of 6798 kWh/yr and dividing it by 10205 kWh/yr, you get an estimated derating factor of 0.67. Again, said another way, your system is 67% efficient across an entire year, 10 points below default.

In summary, you're going to need to understand the basis for the 7858 kWh/yr estimate provided to you via 3 different contract proposals. Either all 3 contract proposals made the same error, and your system's working fine at 2 points better than default 77% efficiency, or, you've got something wrong with your system to cause it to be at 10 points worse than default efficiency.

Hope this helps.

Best regards,

Bill

lesyl

Re: Reasons why my system is underperforming

Ggun – I won’t go into all the “homework” I did, but suffice it to say that I checked into every angle of the proposals and companies that I could think of. I did double check the estimates using a different online calculator. Maybe it wasn’t as good as PVWatts. But it supported their claims.

Mlail and solar-dave – as I mentioned below, I wash the panels regularly and keep them as clean as possible.

Bill – Thanks for the extremely detailed response. I’m going to need to spend more time than I have at the moment going through it.

drees

Re: Reasons why my system is underperforming

a0128958 wrote: »

The 6651 kWh/yr estimate assums dust free panels.

PVwatts does include dirt in it's 0.77 derate factor - 5% loss due to soiling which would be good for light dirt / typical soiling - if they are really dirty then it would not be sufficient. A quick glance at the panels would be sufficient to see if the panels have more than typical soiling.

Calculator for Overall DC to AC Derate Factor

There's a lot of factors that go into system performance. Having a pic of the installed system would help identify issues.

It's also quite possible that a panel is under performing which would drag down the performance of the rest of the system. The only way to help track this is to take periodic performance measurements (energy production over a month, peak system power output on a clear day, total system energy on a clear day) and compare.

a0128958

Re: Reasons why my system is underperforming

drees wrote: »

PVwatts does include dirt in it's 0.77 derate factor - 5% loss due to soiling which would be good for light dirt / typical soiling - if they are really dirty then it would not be sufficient. A quick glance at the panels would be sufficient to see if the panels have more than typical soiling. ...

Thank you! Learned something new. I took out references to dust in my earlier posting since a reasonable amount is now covered by the default 0.77 derate factor.

Best regards,

Bill

ggunn

Re: Reasons why my system is underperforming

a0128958 wrote: »

It looks like the heat in Tuscon significantly affects the output.

Using the same PVWatts data noted above, average daily solar hours in Tuscon is 6.59 (25% more than what we have here in Dallas, BTW).

Multiply this times 365 days/yr and multiply time 4.14 kW, you get an estimated 9958 kWh/year. This is the 'nameplate' estimate for a 4.14 kW DC system in Tuscon. It takes into account all factors that can influence insolation (clouds, humidity, haze, time of year, distance to the sun, etc) for Tuscon, but, does not take into account temperature (in fact it assumes STC temp). It also doesn't include dust sitting on the panels.

You have managed to confuse me, which some will say is not that hard to do... ;^)

From this data:
http://rredc.nrel.gov/solar/old_data/nsrdb/1961-1990/redbook/sum2/23160.txt

I can see that the daily insolation on a flat surface, latitude tilted, azimuth 180 degrees, in Tucson averages 6.5 kWh/m^2/day over a year's time. That is how much energy from the sun strikes the surface in the course of a day, averaged over a year and normalized to our old friend 1000W/m^2.

True enough, multiplying that 6.5kWh/m^2/day by 365 days/year yields 2372.5 kWh/m^2/year (the total cumulative solar insolation on that surface for a year), but I do not see how multiplying that insolation number by the DC rating of the system yields a number that makes any sense. In fact, when I do that calculation and preserve the orders of magnitude (never mind the dimensional units), the raw number I get is 9,822,150,000. You can't just ignore those k's; they mean X1000, and the units you end up with must be Watt-hours or kilowatt-hours.

Either I misunderstood you somewhere or one of us has got this wrong.

a0128958

Re: Reasons why my system is underperforming

ggunn wrote: »

... Either I misunderstood you somewhere or one of us has got this wrong.

One of us is wrong! I look forward to learning here from those smarter than me.

I'll go first to try to clarify:

From PVWatts Ver 1 for Tuscon, AZ: 6.59 kWh/m²/day insolation estimate

6.59 kWh/m²/day / 1000 kW/m² = 6.59 hrs/day of calibrated sunshine since a solar panel is tested at an insolation of 1 kW/m².

6.59 hrs/day * 365 days/yr = 2405 hrs/yr of calibrated sunlight

System is 18 panels rated at 230 Watts (STC) each

18 panels * 230 Watts/panel / 1000 Watts/kW = 4.14 kW system rating

4.14 kW * 2405 hrs/yr = 9958 kWh/yr nameplate estimate

9958 kWh/yr matches the number I calculated above in my earlier posting. This 'nameplate' estimate is commonly used here in the Dallas area as part of the payback investment and/or ROI analysis portion of sales presentations. Unfortunately it reflects each panel's power at too high of a number versus real world output.

Hope this clarifies the calculation.

Best regards,

Bill

Lee Dodge

Re: Reasons why my system is underperforming

In a bit of a deviation from the point of this thread, I have a question about a comment made by some of you. You mention washing your panels regularly to remove the dust. I do not wash my panels for fear of leaving hard water spots on the glass surfaces. I noticed in the last house that I lived in, my windows were very hard to clean on the outside after a number of years. I concluded that the spots were hard water stains left by the sprinkler hitting the windows, as the number of spots corresponded to how much they were sprinkled. The city water was very high in hardness (CaCO3 I believe).

Now rainwater is always soft, while ground water can be hard, so I hope for rain to clean them without leaving water spots! How do you guys avoid hard water spots? Do you squeegee them, which would be effective but add work? I thought about using deionized water, but thats a bunch of water.

Back on topic, I have compared three systems here in southern Colorado with predictions by PVWatts V.2, and these systems inexplicably exceed the predictions:
http://www.residentialenergylaboratory.com/comparison_of_pv_systems.html

It looks like in lesyl's case, the PVWatts predictions are close to what he is producing. It appears that the installers are fudging their estimates, but lesyl should probably have done the PVWatts calculations himself before making such an investment.

ggunn

Re: Reasons why my system is underperforming

a0128958 wrote: »

One of us is wrong! I look forward to learning here from those smarter than me.

I'll go first to try to clarify:

6.59 kWh/m²/day = 6.59 hrs of calibrated sunshine since a solar panel is tested at an insolation of 1 kW/m².

6.59 hrs/day * 365 days/yr = 2405 hrs/yr of sunlight

4.14 kW system rating * 2405 hrs/yr = 9958 kWh/yr

9958 kWh/yr matches the number I calculated above in my earlier posting.

Hope this clarifies the calculation.

Best regards,

Bill

I don't think so. That number is the cumulative amount of energy that is projected onto a surface at that tilt and azimuth by the sun over the course of an average day (averaged over a year's time), normalized to 1000Watts/m^2 perpendicular to the surface at solar noon. It has nothing to do with how much energy a solar module will harvest from it, which would be somewhere around 14-16% of that number in an average day (in kWh) if the module were crystalline silicon and 1 square meter in size.

When you are making these calculations, you have to carry the k's through as X1000, so if you have one term with a k in it multiplied by another with a k in it, the product has to have an M (X1,000,000) in it.

Also, the other dimensional terms must work out as well, i.e., (kWh/m^2/day) multiplied by (days/year) yields (kWh/m^2/year) because the units combine multiplicatively and the "day" term cancels out . (kWh/m^2/year) multiplied by (kW) makes no sense; the units would be ((kW)^2)h/m^2/year.

a0128958

Re: Reasons why my system is underperforming

Here's where I learned this from:

http://www.microcircuitlabs.com/spb.htm

See end of article.

A panel's power rating is based on STC measurement at an illumination of 1 kW/m².

Best regards,

Bill

stephendv

Re: Reasons why my system is underperforming

Probably by sheer luck, using the pvgis estimation tool http://re.jrc.ec.europa.eu/pvgis/apps4/pvest.php?map=africa and picking a spot in Morocco at 35.1 degrees latitude, with all the default values, the yearly estimate comes out to 6570kWh/year.

ggunn

Re: Reasons why my system is underperforming

a0128958 wrote: »

Here's where I learned this from:

http://www.microcircuitlabs.com/spb.htm

See end of article.

A panel's power rating is based on STC measurement at an illumination of 1 kW/m².

Best regards,

Bill

Ah, I see the problem with the equation you posted. The term you left out was the divide by Io at the end, which is 1kW/m^2. Although looks like unity, it's a divide by 1000 that cancels out one of the k's in the numerator and the units are kW/m^2 which makes the result come out in kWh. I am still reading through the article to figure out what they are actually saying (I'm not sure I agree with it), but that's why your equation looked bogus to me.

a0128958

Re: Reasons why my system is underperforming

ggunn wrote: »

Ah, I see the problem with the equation you posted. The term you left out was the divide by Io at the end, which is 1kW/m^2. Although looks like unity, it's a divide by 1000 that cancels out one of the k's in the numerator and the units are kW/m^2 which makes the result come out in kWh. I am still reading through the article to figure out what they are actually saying (I'm not sure I agree with it), but that's why your equation looked bogus to me.

OK. I agree. I updated the calculation analysis above that I wrote yesterday to add the term I left out. I also made some minor changes to make sure the units are consistently carried through from line to line.

I look forward to learning anything you come up with. I'll fix the calculation analysis further if need be so that future readers will see it stated correctly.

Best regards,

Bill

ggunn

Re: Reasons why my system is underperforming

a0128958 wrote: »

OK. I agree. I updated the calculation analysis above that I wrote yesterday to add the term I left out. I also made some minor changes to make sure the units are consistently carried through from line to line.

I look forward to learning anything you come up with. I'll fix the calculation analysis further if need be so that future readers will see it stated correctly.

Best regards,

Bill

I am still examining the article, but it looks like their projections are based on performance at STC (standard test conditions), which as we all know do not exist in the real world. STC measurements are designed to provide an apples to apples comparison between modules to assess their performance relative to each other, not to estimate how they will perform in the field. Even if their numbers are correctly derived (which I still am not sure of), I don't know that they would have any real world significance. Why should we care what a system would produce if the cell temperature were always 25 degrees C?

ggunn

Re: Reasons why my system is underperforming

ggunn wrote: »

I am still examining the article, but it looks like their projections are based on performance at STC (standard test conditions), which as we all know do not exist in the real world. STC measurements are designed to provide an apples to apples comparison between modules to assess their performance relative to each other, not to estimate how they will perform in the field. Even if their numbers are correctly derived (which I still am not sure of), I don't know that they would have any real world significance. Why should we care what a system would produce if the cell temperature were always 25 degrees C?

OK, I think I've got it. The calculations are OK as far as they go, but you have taken them out of context. Their calculations are meant to show that if their gizmo is at the same location and orientation as the PV array, it can project what the array should be producing based on a simple scaling algorithm since they know what their little onboard panel produces and it is subject to the same conditions as the larger array, i.e., the deviations from STC cancel out.

Standing alone, the computations are meaningless.

a0128958

Re: Reasons why my system is underperforming

ggunn wrote: »

... The calculations are OK as far as they go, but you have taken them out of context...

Thanks for the comments.

Manufacturers of panels commonly quote the panel's power at STC. For example, the panels being discussed here are commonly refered to as 230 W panels. We know in fact that in practicality, they don't produce 230 W - it varies with respect many factors.

Similarly to panel manufacturers, some installers here in the Dallas area use the customer's location's insolation (which we know is stated at STC) along with the panel's STC power, to quote an annual 'harvest' output of the system. I think we're all agreed that this is an STC based 'harvest' and not what actuality will be for a 365 day kWh total.

When lesyl said his system isn't going to produce the kWh/yr that matches the higher estimate he received from 3 different proposals, I did the 'nameplate' system capacity calculation thinking initially that this was the problem. I.e., lesyl was quoted 'nameplate system capacity, similarly to what's done commonly in the Dallas area.

But, the 'nameplate' capacity of lesyl's system is much higher than the estimate amount that was actually provided to him/her. The calculation was useful to show this.

Best regards,

Bill

tallgirl

Re: Reasons why my system is underperforming

mlail wrote: »

As someone pointed out, dust is a problem over the long run. I know the Pheonix area was hit with dust storms. As I understand it, your area is not as prone this this but without a regular rain to clean off the panels, you might want to clean them manually.

Dust is nothing compared to heat -- we had a drought and I intentionally didn't wash my arrays. The temperature-corrected drop in performance was around 3%. The temperature-related drop in performance was something like 15%.

Proper instrumentation is your friend.

ggunn

Re: Reasons why my system is underperforming

a0128958 wrote: »

Thanks for the comments.

Manufacturers of panels commonly quote the panel's power at STC. For example, the panels being discussed here are commonly refered to as 230 W panels. We know in fact that in practicality, they don't produce 230 W - it varies with respect many factors.

Similarly to panel manufacturers, some installers here in the Dallas area use the customer's location's insolation (which we know is stated at STC) along with the panel's STC power, to quote an annual 'harvest' output of the system. I think we're all agreed that this is an STC based 'harvest' and not what actuality will be for a 365 day kWh total.

When lesyl said his system isn't going to produce the kWh/yr that matches the higher estimate he received from 3 different proposals, I did the 'nameplate' system capacity calculation thinking initially that this was the problem. I.e., lesyl was quoted 'nameplate system capacity, similarly to what's done commonly in the Dallas area.

But, the 'nameplate' capacity of lesyl's system is much higher than the estimate amount that was actually provided to him/her. The calculation was useful to show this.

Best regards,

Bill

If that's what integrators are telling their prospective customers, then it's a shady (pun intended) practice. This is the kind of thing that gives the solar industry a black eye. It reminds me of Brad Pitt punching himself in the face in Fight Club.

I would never pitch a system to a customer based on such bogus numbers. Overpromising performance in order to sell a system is very bad business. Even in Dallas. ;^)

lesyl

Re: Reasons why my system is underperforming

Here's the latest short version of a long story: The owner of the company has agreed that the system is underperforming and that it should have produced around 600kWh more to date than it has. After seeing a photograph of the system at 9 a.m. and a site visit, he believes the problem is the electric service mast that partially shades two panels in the morning. He said their tech (who spent about two hours going through the system in the middle of the day) found the system performing normally otherwise. He gave me a detailed spreadsheet so I could track production for at least a month. Based on those results, they’ll either add a panel or two or move two panels. He contacted the electric company to see if the mast could be lowered, but they apparently said no.

The obvious question of course is shouldn’t they have known the mast would be problem. I haven’t asked it yet, but I will. I have a vague recollection of discussing the mast with them. I suppose it’s possible we had some kind of miscommunication. But I was clear that we wanted the best placement and angles to catch as much sun as possible.

That’s where things stand as of today.

drees

Re: Reasons why my system is underperforming

Sounds like they screwed up - they definitely should have caught the shading issue.

Do you have an overhead picture of your house and a picture of the panels w/the mast?

A plot of your system's output (look at the link in my sig for my system's output to compare) on a clear day should clearly show the affect of the shading, too.

Micro-inverters would have avoided this issue (the shading would only affect the shaded panels instead of the entire array) but they do cost a bit more than a central inverter.

a0128958

Re: Reasons why my system is underperforming

lesyl wrote: »

... He gave me a detailed spreadsheet so I could track production for at least a month. Based on those results, they’ll either add a panel or two or move two panels. ...

I don't think you need a month's worth of detailed data to show the problem. The first day that is cloudless should suffice.

Here's an example. By plotting the solar PV system's output on a cloudless day as a function of time (below), it's clear to see that the shape of the output is 'stunted' in the morning. At this time, by observation, it can be seen (also below) where the shading is (chimney affecting 2 of 12 panels in a string). You could do this all in one day, once you've got the right kind of day (cloudless).


Best regards,

Bill

Blackcherry04

Re: Reasons why my system is underperforming

Here is a Enphase install, of the 3 shaded panels , one is a furnace vent, one is a plumbing vent that is only 1 " above the panel and the other is a combiner box that is 4 feet from the array, I never thought would cause a issue. It sure doesn't take much to be a problem. Over 25 years, it could be a big number.