One of the most critical parts of the Financial Model of a PV plant is the accurate estimation of the capital expenses (CAPEX). The aggregated value of CAPEX is fundamentally composed of two costs groups: the main equipment (PV modules, which are highly volatile and sensitive to Chinese market demand, and inverters and racking systems, which are much more stable and predictable), and the balance of system (BOS), including among other secondary materials (cabling, combiner boxes, etc.), installation works (civil, electrical, mechanical) and management costs.
However, the total USD/Wp costs of the BOS — and, hence, the relative weight of each of those costs groups within the overall CAPEX figure — vary substantially depending on the DC capacity installed, considering we compare plants with identical technology (overall design, modules, racking, inverters, etc.). For instance: For a PV plant with mono-PERC modules and single-axis trackers, the weight-ratio BOS versus main equipment might vary from roughly 25%/75% for a 100MWp PV plant to 50%/50% for a 100MWp PV plant, assuming the exact same unitary prices of main equipment for both.
The Relevance Of Indirect Costs
Undoubtedly, the most relevant item within the BOS costs when it comes to variation with DC capacity is the indirect costs. These include all items of an EPC contractor that are not directly related to PV plant materials supply and installation, such as project management costs from main contractor and subcontractors, mobilization, secondary tasks during construction, contingencies (often used to protect against potential liquidated damages (LD) or construction risks) and markups to supplies and services.
At first sight, it could seem reasonable to think that building a 3MWp PV plant would entail 33 times less cost than a 100MWp PV plant because we mainly think about the supplies and installation work required. However, this ignores the relevance of indirect costs and two important elements:
Construction time: This is indirectly related to the DC capacity, with a much higher time/MWp ratio for smaller plants. This is due to several factors, including like the much lower efficiency of resources in smaller PV plants and the manufacturing and shipping time for main equipment (e.g., for inverters and transformer stations, it’s rarely less than 18 weeks). This directly affects all expenses linked to the duration of the construction and the abovementioned secondary tasks during construction, such as day/night presential site security, health and safety (H&S), quality assurance (QA) or waste management.
Management of resources: An 1MWp PV plant, for instance, requires one project manager and one site manager. The same resources would serve to manage a 30MWp (or even larger) site, and only one more site manager would need to be added for a 100MWp one.
Watch The Perimeter
The perimeter of the PV plant — often mistaken as proportional to the area of the plant — is another factor that’s usually underestimated. The perimeter is not increasing proportionally with area. If you assume, for instance, a perfectly square plot – leaving aside the impact of irregular perimeters — an area of 100m2 would have a perimeter of 40m, whereas an area of 400m2 (four times larger) would have a perimeter of only 80m (two times larger).
In other words, in a PV plant with four times more capacity, we would be spending only two times more on all items related to the perimeter, such as fencing supply and installation, perimeter security system (e.g., CCTV) and security trenches.
Finally, we need to mention other two costs groups that, despite being obvious, are often undervalued when it comes to BOS estimation for small PV plants.
Interconnection costs: Whereas their impact is obvious, and they’re often a critical point of the development stage, these costs are usually underestimated for medium and small PV plants (<40-50MWp), especially if a high voltage (HV) interconnection is required. Since it depends on the connection voltage (MV -11 to 34.5kV or HV -45 to 400kV) and the connection line type (underground or overhead) and distance (few m to tenths of km), the cost can vary significantly, and there is no way to provide narrow cost ranges. However, we can assume that connecting a medium-size PV plant (e.g., 20MWp) in HV could easily cost more than 0.10 USD/Wp even with a short connection line, while the same interconnection type for a large PV plant (e.g., 100MWp), even with a much larger MV/HV transformer, would only be 0.025 USD/Wp.
Monitoring System: A standard monitoring system would have some components that escalate with the plant size (e.g., the distributed programmable logic controllers (PLC) for each inverter station or the fiber optic) and others that don’t (e.g., the supervisory control and data acquisition control system, whether local or web, or the power plant controller). Considering this, the monitoring system costs could vary from 0.0025 USD/Wp for a 100MWp PV plant to 0.03 USD/Wp for a 3MWp PV plant.
To sum up, the total BOS costs in USD/Wp increases exponentially with the reduction of the DC size, with the curve starting to flatten only beyond ~40-50MWp. It is of utmost importance, therefore, to not underestimate the CAPEX for small and medium-size PV plants.
POST WRITTEN BY
EVP, Global Head of Engineering and Construction, Sonnedix, overseeing the design and construction of assets for global solar PV platform.
As seen on Forbes