This calculator determines the cost to install a PV system based on the properties of the PV modules and the balance of systems (BOS) costs.

It is common to compare PV modules in terms of dollars per peak watt ($/W_p); that is, by the price paid for a module divided by the power it produces under standard measurement conditions. The benefit of this metric is that it is simple, and that it helps to compare modules of different sizes and efficiencies.

The problem with comparing modules in $/W_p, however, is that PV electricity is not generated by modules alone. A PV installation also requires land, frames, labour, wiring, inverters, and other electronics, all of which are lumped under the term, Balance of Systems (BOS). These BOS costs depend on the efficiency and the structure of the PV modules.

In general, comparing the price of PV modules on a $/W_p basis makes higher efficient modules appear less cost-effective than they really are.

For example, if one could purchase a 10% module for $100, or 20% module for $200, and both modules had the same area of 1 m², then on a $/W_p basis, the modules would cost the same: 1 $/W_p (see equations below). Yet this does not mean that a PV installation using the 10% modules would cost the same as one using the 20% modules.

Consider the installation of a 20 kW_p PV power plant. One could either install 100 m² of the 20% modules or 200 m² of the 10% modules. It would clearly be cheaper to use the 20% modules, since that would require less land, less labour, and fewer frames, all of which make up a significant fraction of the cost of an installed system.

There can also be structural differences that make one panel superior to another, but which are not entailed in the $/W_p metric. For example, lighter panels might incur lower transportation and framing costs, annd flexible panels might be quicker to install.

BOS costs vary significantly from one application to another, and from one country to another, so they cannot be evaluated with a generic value or fraction. Nevertheless, BOS costs should not be neglected because they typically contribute to more than half of the cost to install a PV system.

It is pertinent to note here that the value of a PV module does not affect the installation cost alone. Different types of modules also repond differently to changing operating conditions (temperature, spectra, intensity), they degrade at different rates, they have different failure modes, they require different degrees of maintenance, and so forth. These operational aspects are not encompassed by this calculator, which is specific to installed PV costs. Thus, the installed PV cost is one step superior to comparing module prices in terms of $/W_p, but more analysis is required to give a complete assessment of a module's value to a PV installation.

The installed system cost C_sys is given by:

C_sys   =   N_mod ⋅ C_mod   +   C_BOS,

where N_mod is the number of installed modules, C_mod is the cost of each module (assuming all modules are identical), and C_BOS is the cost of the balance of systems.

The number of modules N_mod is equal to the system power P_sys divided by the module power P_mod, where the latter depends on the efficiency η_mod and area A_mod of each PV module:

N_mod   =   P_sys / P_mod   =    P_sys / ( η_mod ⋅ A_mod ⋅ 1 kW_p/m²).

The balance of systems cost C_BOS is rather complicated to calculate as it is comprised of many factors: e.g., labour, frames, cabling, inverters, insurance, transportation, land. In this calculator, they are simply represented by four variables:

C_M, BOS costs that scale with the number of modules in $/module, like the cost of bolting down and wiring up each module.

C_A, BOS costs that scale with area in $/m², like the land or roof area.

C_P, BOS costs that scale with system power in $/W_p. This includes aspects of the inverter since it typically becomes more expensive to transform more DC power into AC power. Inverter costs are indpendent of the system area.

C_F, BOS costs that are fixed and independent of the system size or power. These can include some administration costs and other aspects to labour, the inverter, and cabling between modules and inverter.

Hence, the BOS costs can be represented by the equation,

C_BOS   =   N_mod ⋅ C_M   +    N_mod ⋅ A_mod ⋅ C_A   +    P_sys ⋅ C_P   +   C_F.

In 2020, we added a new equation to account for how the price of the modules C_mod depends on module efficiency η. The equation was proposed by Paul Basore in his white paper on LCOE, where

C_mod(η)   =   C_mod(η_ref) ⋅ { α + (1 – α) ⋅ (η / η_ref) ^ [ln(2) / ln(β)] },

where C_mod(η_ref) is the price of the modules at the reference efficiency η_ref, where reference price and efficiency are entered under 'user inputs'; and where appropriate values for α and β are 0.5 and 1.2. This equation provides a lower limit to module price (i.e., a non-zero price when η = 0%) while ensuring that module price increases as cell efficiency increases (primarily because higher efficiency cells require more expensive fabrication methods). The equation will make sense once you plot it.

Although it is well known that a comparison of modules prices on a $/W_p basis makes a high efficiency module appear less cost-effective than it really is, such comparisons are widely used because (i) they are simple, and (ii) there is not a set of BOS costs generally accepted by the PV industry as being representative for a particular application. With this calculator, we would like to list a series of case studies that can be used for this purpose.

Case studies that break down BOS costs are rarely published because they are confidential to installation companies.

One dated case study published in 2010 was specific to 1.5 kW_p residential rooftop systems installed in Canberra, Australia, in 2010 [1]. It concluded that C_A = 190 +/– 50 AUD/m², C_P = 0.62 +/– 0.10 AUD/W_p, and C_F = 1500 +/– 400 AUD. C_A is high for small roof-top systems such as these because of the significant labour associated with their installation; in this study, C_A is compounded by the high labour costs prevalent in Australia.

Another more recent case study for a 5.76 kW_p residential rooftop system (20 modules) installed in the USA in 2020 [2] gives C_M = 40 USD/module, C_A = 95 USD/m², C_P = 0.15 USD/W_p, C_F = 3500 USD, α = 0.5, β = 1.2, η = 18%, A_mod = 1.6 m². These inputs are currently the defaults.

You're welcome to send other published case studies to support@pvlighthouse.com.au

[1]	T. Brazier and K.R. McIntosh, "The dependence of the installed cost of a 1.5 kW rooftop PV system on module efficiency," Proc. 48th AuSES conference, Canberra, 2010.
[2]	P. Basore, "Levelised cost of electricity (LCOE)," White paper posted on PV Lighthouse, June 2020.

Please email corrections, comments or suggestions to support@pvlighthouse.com.au.