Why collaboration is the crucial factor in designing the most efficient PV farm layout
Updated: Sep 17, 2022
Collaboration is the action of working with someone to produce something (similar: cooperation, alliance, teamwork, joint effort, working together)
What is PV Farm Layout?
The solar PV farm layout is a racking/mounting plan of PV modules which shows the information about the site, POI and designed system characteristics. PV farm layout is the most important document that reflects chosen single-line diagram and equipment specification and serves as a basis for energy performance analysis, cost estimation, construction schedule and construction drawings.
During the project lifecycle, the layout is updated dozens and hundreds of times; it is like the place where all disciplines come together and connect, from electrical and civil engineering to procurement, staging, logistics and many more.
Think about the solar farm layouts as floor plans for single-family houses.
Project size and team size
Solar farms (as their layouts) can be of different sizes - from a couple of megawatts to thousands of megawatts or from a couple of acres to thousands of acres.
Collaboration is about people, so let's discuss how many people are involved in solar PV farm projects during the design and preconstruction phases. For example, a small 1MW project can be fully designed by a couple of people, while one around 100MW will require dozens of people.
Although the project size and team working on it do not depend linearly, it is fair to say that the more significant project requires more people to be involved in the design and preconstruction.
As you see from the chart above, the number of people involved in the design and preconstruction snowballs from 5 MW to 50 MW; after 50 MW, it slows down, growing little by little. Thus the number of people involved in the design and preconstruction of, say, 330 MW project and 470 MW project will be approximately the same.
PV farm layouts are plans that contain many repeatable parts. So the project size and complexity do not depend on each other linearly. For example, the solar PV farm 330 MW project can be much more complex than the 470 MW one because of geology conditions or sophisticated boundaries (with the same number of people working on both projects).
Team size and communication complexity
The communication complexity grows quadratically with the team size (Metcalfe's law), but the consequences of broken communication grow exponentially. Therefore, in designing PV farms or developing software products, broken communication is one of the biggest threats to the project.
Multidisciplinary process flow and feedback loops
Let's go back to PV farm layouts.
We realised that the more significant project we have, the more people are involved in its design and preconstruction. The job those people do is broken down into areas of expertise or disciplines.
Many areas of expertise are overlapped when producing the layout. Here are the most important ones: energy performance modelling, civil engineering, electrical engineering, procurement engineering, staging and logistics, and construction (sequencing, work packaging, labour operations, etc.).
Here is an example of the process flow for an average utility-scale PV farm project (Owner/Developer - EPC Contractor):
The Owner/Developer went through a feasibility study and found funding (loan) to build a PV farm.
The layout created for this study's sake was a conceptual one built on assumptions about RLV, Gross Energy Value (PPA) and EPC Contract Cost.
The site for the solar PV farm is financially viable if RLV = GEV - CC is good enough to invest in the project.
So the Owner/Developer creates RFP and organises a tender communicating with EPC contractors about site details and layout specifics.
The main point of those conversations is to find ways to be within assumptions the owner made when doing RLV calculations or change them but still make the project financially viable.
The communication between those different areas of expertise creates feedback loops.
What is a feedback loop?
A feedback loop is a process in which the outputs of a system are circled back to adjust the input until all process participants find an agreement. For example, the solar design feedback loop refers to analysing PV farm layout metrics (the outputs of a service or product) to create a better layout.
Regarding the process flow we considered above, there are external feedback loops (Owner/Developer reps and EPC Contractor reps) and internal ones (between discipline reps inside EPC Contractor).
External feedback loop example
Plan Goal (Developer)
Can we get a 516,840 MWh/year by using LONGI 550 mods and 70 Sungrow 3600 inverters (295 MWdc) on the site below?
Design Scenario (EPC Contractor)
Yes, we can:
Analyse trade-off (Developer and EPC Contractor)
D: The cost for the EPC Contract is too high; it increases our cost per Watt DC
C: To achieve your spec requirements, we needed to go tight: shift trackers and use small (1 string, 29 modules) ones with a pitch of 20 feet and 2-height preference
D: Is there any way to reduce the construction cost?
C: Could you specify it from the very beginning?
D: Okay, let's start from the very beginning.
As we see from the above, loop #0 was a warming-up. The whole point was to understand better initial requirements or to showcase why certain things don't work. So the inputs (goals and what-ifs) will be adjusted for the second iteration.
Plan What ifs
What if we increase the row-to-row space, trying to keep the same first-year energy but reducing the capacity?
What if we use only 116 mods (4-string) trackers, keeping bigger row-to-row space?
What if we use only 116 mods (4-strings) trackers and apply BLA (trunk bus) electrical pattern?
What if #1
What if #2
What if #3
As we see from the numbers above, option #1 gave us less DC capacity than the base case, but at the same time, it gave us more energy, which means that by increasing row-to-row space, we reduced near shading (from neighbouring trackers). In addition, the first option costs less than the base case. It might be something the Owner/Developer didn't consider when doing the feasibility study. To understand whether those numbers can work in their spreadsheet, they need to run this scenario through a financial assessment.
Further scenario refinement led us to reduce DC capacity and energy compared to scenario #1. And again, to compare scenarios #1 and #3 is not enough to summarise something like that "scenario #3 gives us $51.1 million less in over 25 years (~$2 million per year), but the construction cost in the next one year will be $21.2 million cheaper". Cost numbers need to go into the Owner/Developer's spreadsheet to compare scenarios properly, where together with PPA, Investment, Debt Contribution, Cash Flow, and IRR, they will start dancing nicely (or not).
Internal feedback loops examples
What if we design only perfectly balanced blocks? How much capacity will we lose compared to the 'maximum power' option?
The starting point of working on the layout is to get the maximum possible power to start understanding the boundaries. This happens with less attention to balancing or grouping trackers to the blocks (connecting them to combiner boxes, inverters and transformers).
Let's compare those two options:
Revenue 25 years, $M
As we see in the table above, even though 'maximum' power will bring you more energy and revenue over 25 years, it is not the best option. The 'best balance' option will generate less power, but at the same time, it will be much cheaper to build (fewer inverters, fewer modules, fewer trackers, less wiring, less labour).
In addition, although we have 73 inverters vs 64 and the average DC/AC ratio of 1.23 vs 1.21, the total deviation is much higher for the 'maximum power' option - there are clipped power inverters as well as unloaded ones, which makes it even less efficient from an energy standpoint.
And that is not taking into account operation and maintenance costs which also will be lower than for the 'maximum power scheme'.
What if we compare the cost savings of ensuring an average 1.0% DC loss to a 1.5% DC loss?
DC circuits shall be designed to have a maximum loss of 1.5% at STC with an average loss of 1.2% at STC
The above is quite a common requirement from Owners/Developers. People think keeping the lowest possible voltage drop is the best for energy production (which is true), but sometimes the cost is not worth it. This type of what-if analysis is expected, so let's see what we get.
Let's compare two options from the above:
As we see, the cost of reducing losses from 1.5% to 1.0% would be ~ $4.8 million. While the increased energy revenue will be only ~ $2.3 million over a lifetime. Spending ~$4.8 million to make $2.3 million in 25 years is a bad idea.
What if we move array A to place B? What would be the difference between civil and electrical costs?
That is a typical what-if people don't usually have enough time to validate. Some areas are closer to the substation, so the MV wiring is cheaper, but they need much more dirt to move to install equipment, while other areas are the opposite. This leads us to comparing wire feet and dirt cubic yards (or classical apples and pears).
So what if we move array A to place B?
Let's compare the cost to build.
Equipment cost, $M
Electrical cost, $M
Mechanical cost, $M
Civil cost, $M
Total cost, $M
As we see in the table above, moving blocks from area A to area B reduced electrical costs by $7.2 million while increasing civil costs only by $3 million. So it might be an excellent business decision to make.
Why collaboration is the crucial factor in designing the most efficient PV farm layout
First, let's clarify some things: 1) collaboration involves working with someone to produce something, and 2) collaboration presupposes communication.
If we will take our feedback loop and cut it in a straight line into two parts: something that we can do only collaboratively and something that we can do by solo effort, we will get the scheme below.
We can only do activities like planning goals and analysing trade-offs by working collaboratively. This part of the feedback loop is the most significant one.
At the same time, activities like designing scenarios not always can be done by solo effort, sometimes they require collaborative work too. This means that collaborative work is an even bigger piece than what we see in the scheme above.
To add a bit more weight in addition to the Feedback Loop mental model, I would mention that we as a company have been involved in hundreds of conversations about feasibility studies around residential, mixed-use buildings and solar farms. Also, we have been shadowing design and preconstruction meetings at different phases of the PV farm project.
80% of the time people spend on collaboration (synchronisation and communication) and 20% on making solo efforts (it might vary from project to project or from phase to phase). So the first reason why collaboration is the crucial factor in designing the most efficient PV farm layout is connected with the fact that in a typical design process with multiple project shareholders involved, it simply has a more enormous mass in comparison to a solo effort.
The second reason is related to the main goal of the feedback loop - analysing PV farm layouts to find the most efficient one.
Collaboration is a chain of the whole feedback mechanism, and a broken chain prevents us from finding sought-for the most efficient PV farm layout.
By the way, we can find an exact definition of 'the most efficient one' only through collaboration; otherwise, civil engineers will think about minimising grading, while electrical engineers will push for the cheapest wiring. I bet we will get different layouts.
Okay, collaboration is essential, so what?
If we are industry players (developers or contractors), we should invest in collaboration and spend time improving it. No technology can save the team if they don't work together.
If we are software providers for designing buildings and solar farms, we should build products with collaboration in mind.