Wake up and smell the system

Issue 5 and Volume 101.

Wake up and smell the system

Business as usual rarely leads to significant breakthroughs in technology. At last September?s Photovoltaic (PV) Systems Performance and Reliability Workshop hosted by the National Renewable Energy Laboratory, two Sandia National Laboratory scientists sent a wake-up call to the PV industry, urging industry players to seriously question decade-old assumptions about system design.

Mike Thomas and Hal Post made it clear in their workshop article, OPhotovoltaic Systems Performance and Reliability: Myths, Facts and Concerns?A 1996 Perspective,O that Ocontinuing to follow the approach of the past, which is focused on individual components, will encourage improvements in components that may be of minimal benefit to today?s marketplace.O

According to the authors, one reason the trend toward component-centric research is so prevalent is the relationship between system pricing and module advances is not widely understood. In grid-tied systems, for example, the PV modules represent about 50 percent of the overall costs.

For stand-alone or hybrid systems which include energy storage, the module costs drop to between 25 and 30 percent of system costs. So a 50 percent decrease in module cost may translate to as little as 12.5 percent savings in the overall picture. Likewise, increasing the module efficiency from 13 to 14 percent would, at most, result in a 4 percent decrease in overall cost in a grid-tied system?much less in a storage system.

Thomas and Post also identified problems with the status quo in system reliability and performance. For example, in a study of hybrid PV-engine systems, the results showed that PV array utilization can be as low as 65 percent. Combine this under-utilization with system efficiencies as low as 50 percent because of losses in inverters, batteries and voltage drops, and an unusual situation develops; for many hybrid systems, most of the energy produced by the renewable energy systems never makes it to the load.

Thomas and Post admit there are still benefits from the engine side of the hybrid equation (reduced run hours, lower fuel consumption), but the economic justifications used for these systems go largely unrealized. OUnfortunately,O said the authors, Omany current designs E are being driven by fuel and engine run-time arguments.O

Batteries and inverters are the largest sources of uncertainty in today?s PV systems. For battery storage systems, the life-cycle costs are 35 percent higher than expected because of battery replacements alone. Yet in the United States, research contracts in these areas focus on long-term R&D rather than improving existing products.

System degradation and maintenance were other areas the industry has overlooked in the name of progress. Consider PV modules, which have the longest warranties in the overall system (up to 20 years with no more than 10 percent degradation in 10 years) and are unquestionably the most reliable component. In the authors? experience, many modules are overrated by 10 to 15 percent to begin with. Now apply the typical degradation rate of 1 to 2 percent per year, and the warranty degradation is quickly surpassed.

However, since few systems measure performance accurately, or at all, warranty claims are minimal. What this means in terms of maintenance is an increase in levelized energy costs of $0.04 to $0.08/kWh. The result, according to Thomas and Post, is Oto divert attention from the need to establish required preventive maintenance activities and to significantly overestimate the value of energy production.O

Still other problems exist in system design. For example, OnetO metering, which allows two-way flow through the revenue meter for grid-tied applications, is advantageous for current system owners. However, net metering ignores the potential benefits of energy storage.

Another design problem is the use of single-unit power processors, regardless of array field size. The result has been to design the system to the requirements of the power conditioning hardware instead of to the needs of the array field. Likewise, power conditioners are chosen to meet instantaneous loads rather than continuous loads. This results in oversized inverters in many systems. Consequentially, energy produced is reduced by as much as half because of inefficient production at low loads. According to the authors, this could possibly be remedied by staging the inverters for high efficiency over the full load range.

Thomas and Post conclude it is clear that the PV community can contribute solutions to these problems. However, Owithout making improvements that benefit the systems of today and tomorrow, investment in component improvements may be of little value, market expansion will be slowed and U.S. industry may be eclipsed by foreign competition.O