Plane of Array Vs Global Incident in Coll. Plane

[Tokoyoshi]

My firm just finished constructed a 17.5 MWP plant a few months ago and we are in the process of analyzing the measured vs PVsyst estimated production data.

Our Scada is giving us the Plane of Array Irradiance (POA) in w/m^2. Which of these PVsyst data is equivalent to the measured POA Irradiance? The "Global Incident in Coll. Plane" or the "Effective Global, Corr. for IAM and Shading"

[André Mermoud]

Your measured value corresponds to the GlobInc value (Incident global), which is just the result of the transposition from horizontal to your tilted plane.

Th GlobEff value (Effective global) is the GlobInc, affected by the optical losses like Far and Linear shadings, IAM and soling losses.

[gonecrawfishin]

It seems to me that GlobEff should be less than GlobHor 100% of the time, because GlobEff includes some losses like IAM and shading. Why would GlobHor be less than GlobEff?

[jromme]

GlobHor is irradiance measured in a horizontal plane. GlobInc is the irradiance that hits the actual orientation of your array. This will likely be greater than GlobHor as there are inherent benefits to tracking or installing modules at an angle. GlobEff is GlobInc after all optical losses are applied, as mentioned below.

For example, a single axis tracking system could see a GlobInc boost of 35% because it's facing the sun most of the day (ideally). After all optical losses are applied, the GlobEff boost could still equal 25% when compared to GlobHor.

[Timothydiz]

The Global incident is just the result of the transposition of your meteo data GlobHor to your tilted plane.

With tracking arrays, when using backtracking, this could change when you change the width of the trackers or the pitch between trackers, because the backtracking conditions will vary.

In any other case, the Incident Global has no relationship in any way with the rest of the system, such for example the AC capacity.

[gonecrawfishin]

Does anybody know how global incident in coll. plane could be negative? That's a new one for me. I just downloaded version 6.7.9, but it was able to successfully run some of my older tracker simulations just fine.

[gonecrawfishin]

Ah, I just answered my own question. In case anybody else has this issue, I had accidentally tried to track from -45deg to -45deg, which admittedly is not a great idea.

[dkourk]

I am new to PVSyst and have a similar problem, not sure if I should open a new thread though.

I have made to simulations for the same site. One has a defined 3d scene, the other does not and uses standard settings. This is a one axis N-S tracking system where I get a significant difference in the GlobInc values (almost 20% lower without the 3d scene) and cannot figure out why. The orientation is optimal N-S in the 3d scene and I see no material losses from shading. The starting value of the Globlinc is materially different and I think this is what produces a very low specific production for the 3d scene simulation. It might be worth noting I am using a different meteo file for the two, however the starting point (GlobHor) is very similar so I do not think this is the cause of the difference.

Any help, guidance or explanation on what I am doing wrong would be much appreciated.

 

tnx, DK

[Michele Oliosi]

I assume you are not using backtracking.

Shading is not relevant for GlobInc, since it has not yet been applied.

Weather can be a factor in so far that every irradiance component has a different transposition scheme. If the pattern of diffuse to direct irradiance is different, there can be differences in the transposed irradiance GlobInc.

Other differences could be found in the details of your tracker definition. Are the stroke limits the same, for example ? Is there any axis tilt ?

[dkourk]

Hi, thanks for the reply, I have tried using backtracking and not. it did not make that much of a difference. Also have tried running the same meteo file and again not a significant difference. 

The difference seems to be coming from introducing a 3d scene but cannot figure out what I am doing wrong or what the parameter of significance is. This is a rather big difference to not understand and could be the difference in the project being viable or not. 

Your help and guidance in figuring this out is much appreciated! I am attaching the 3d scene and orientation settings in case you can spot something wrong as I do not have the knowledge or experience to get to the bottom of this 🙂

tnx, DK

[dkourk]

Maybe the loss diagrams, attached, could help guiding in what is driving this difference?

 

 

[Michele Oliosi]

Ah you have irradiance optimisation on !
Then your movement (therefore GlobInc) could be quite dependent on the weather file. It is therefore crucial to compare apples to apples and start from the same weather file and deactivate the backtracking. Do you have the results for that case ?

[dkourk]

You are right, globlnc now almost matches!

I still have a big difference from what I would expect based on actual nearby data but at least we got to the bottom of this!! 

Any idea what could cause such a difference between expectations based on last 2 years of actual nearby data and simulation?

To be more exact, I have a 0,5MW PV fixed tilt station that is 2kms away from the simulated location and last two years produced an average of 1750Kwh/Kwp/year. I would assume installing one axis N-S trackers would uplift production by about 15%-20%.

Now in the simulation with a couple of different meteo files (solcast TMT and meteonorm 8.1 synthetic) we start from a specific production without tracker of c. 1530 KWh/Kwp and then when having trackers with range between 1687Kwh/jkwp for the 3d scene (a 10% difference vs fixed tilt) and 1799kwh/kwp/year for the generic unlimited trackers simulation.

Overall I am trying to figure what I am doing wrong and cannot simulate more accurately.

- Is it that the meteo files I have do not support the actual data from last two years

- and even so, why does the 3d scene defined simulation only produces a 10% uplift with the trackers?

Any guidance would be much appreciated!!

[Michele Oliosi]

Year-to-year variability can be of about 5%, and climate change can also be a factor. Comparing a couple of years to a TMY may indeed give significantly different results. However, in some cases the satellite TMY / synthetic approaches may be biased up or downwards, that also happens. I would suggest communicating with Solcast and Meteonorm about that.

I am not surprised at the 10% increase, because trackers on a topography are subject to some shadings due to the differences in height, even if you have the best backtracking algorithms.

Note that in PVsyst we are a bit pessimistic regarding the backtracking algorithm. Also note that the irradiance optimization algorithm, instead, is quite optimistic ! Actual trackers may never actually apply such a strategy, there is always a question of lag, rotation speed, actual sky analysis and so on.