System designers have more options today than ever before when architecting solar systems. While this may seem like a great advantage, these options necessitate an ever growing number of decision points in the design process.
No decision has more impact on the system cost and performance than the choice of inverters, as this dictates design constraints for so much of the balance of system. Before selecting brand or model the designer must first choose the macro level class of inverters, central or three phase string inverters. Until recently, the normalized price of string inverters (as measured in $/watt) was much higher than central inverters, making the decision to use central inverters for utility scale quite straightforward.
That unit price gap has greatly diminished, resulting in a heightened debate on the relative merits of central and string inverters, often without empirical data to support the arguments. The jury is still out on which is the so-called “best solution” and likely for good reason, as the overall system size has such a significant impact on the relevant answer.
This article analyzes the relative merits of central and string inverters in a typical system in North America. The analysis is limited to the relative costs of central and string inverters for utility-scale projects in North America in three areas: CAPEX, inverter service life and true cost of service.
The first metric typically used by all designers to make decisions on system architecture is the initial CAPEX cost, and such an analysis is more than a comparison of the relative prices of inverters. It requires a comparison of the costs of the AC and DC balance of system (BOS) and labor as well. Since the costs of modules and racking will be identical these are therefore excluded from this analysis, which compares a 1000V system in North America comprised of 2-MW central inverters to a system built with 60-kW string inverters. The CAPEX analysis is separated into two parts, labor and material.
A full analysis shows that while one 2-MW central inverter skid and associated BOS takes about 19 times as long to install as a single 60-kW string inverter and associated BOS, the fact that there are 33 times as many string inverters to install results in string inverters requiring over 1.7 times the labor to construct a 20-MW system building block. Monetizing this difference at a conservative $40/hr. rate translates to 0.3 cents/watt higher cost of labor installing string-based systems. This ratio would hold exactly in proportion in larger systems above 20 MW without upper bound in system size.
Market experts expect string inverters to be priced 0.5 cents/watt higher than central inverters in North America utility-scale systems. While there is no question that a system using string inverters will have lower DC BOS cost, a complete analysis discloses that these DC BOS savings are more than canceled out by a significant uptick in AC BOS costs, where an important factor is the significant low voltage AC wiring associated with string inverters in comparison to the higher voltage DC wiring of central inverters.
Our analysis shows that a system constructed with 60-kW string inverters will require 1.9 cents/watt more in CAPEX material than a system built with 2-MW central inverters. This is a significant difference in CAPEX. If the 0.3 cents/watt increase for labor is added to this 1.9 cents/watt for material, then string inverters require 2.2 cents/watt more in CAPEX in utility-scale systems.
Inverter Service Life
Utility-scale PV plants are generation assets that are expected to provide a financial return over at least 20 years. A PV plant must therefore operate predictably over that period of time and central inverters have a service life of 20 years or more. On the other hand, string inverters are typically replaced when they fail and there is a risk of a lack of form/fit replacements in the future that could add 25 – 58 cents/watt to the system cost in the event that custom power electronics and electrical retrofit solutions are required. Service life of the inverters is therefore a critical consideration in selection of string or central inverters.
True Cost of Service
There is a common misconception that the cost of service in a utility-scale system built with string inverters is negligible. The only way to completely eliminate the risk in service costs is to obtain a combined extended warranty and service contract from a bankable supplier that fixes future service costs to a specific dollar amount. If service costs of string inverters are compared to central inverters with such a contract, string inverters require an additional 2.6 – 3.7 cents/watt in service costs over a 20-year system life, factoring in their replacement cost and associated labor and truck roll expenses.
While three phase string inverters clearly have their place in the market, this analysis shows that central inverters are the choice for typical utility scale projects in North America, with significant measurable advantages over string inverters in every aspect of the analysis. In addition to the CAPEX and OPEX savings, central inverters have advantages in meeting stringent grid code compliance requirements and end-to-end efficiency benefits. String inverters become much more viable in the face of certain project challenges, such as poor access roads, lack of skilled service teams in the region and non-uniform mismatch due to uneven terrain or shading.
An in-depth version of this article that includes further analysis of utility-scale system performance requirements and operating efficiency as well as considerably more in depth analysis of CAPEX and true cost of service, was originally published by Schneider Electric and this condensed version was republished with permission.