Photovoltaic Technologies – Past, Present and Future
The post that follows is the first part of a two-part paper. The second part will be posted tomorrow.
Photovoltaic technologies' (PVs') decreasing cost and increasing reliability have made this the most attractive generation option for many utilities and facilities. PVs are the most scalable generation option, being cost-effective on many scales ranging from residences to utility-scale projects with capacities that rival the largest cogeneration plants. Rapid expansion in residential deployments in past years helped to create the mass-produced products required for a wide-range of larger projects.
Overall, U.S. new PV installations grew 95% in 2016, and became the largest newly installed source of electric generation (at 39%). Utility-scale installations in 2016 were 10,600 MWdc. In 2016 non-residential (commercial and industrial) installations were 1,600 MWdc and residential installations were 2,600 MWdc.
The pricing of PV project components have been driven down rapidly (see the chart below). The Department of Energy’s SunShot Initiative hopes to reduce the cost of PV-generated electricity by about 75% between 2010 and 2020. In September 2017 SunShot had achieved the utility-scale target.
Historical, current, and SunShot 2020 target system prices for the utility, commercial, and residential sectors (weighted national average for fixed-tilt systems)
SunShot has now set goals for 2030, as can be seen from the chart below.
2.Past and Present
The current success that PV is enjoying can be mostly attributed to production efficiencies in solar panels that have increased over the last decade. These panels are the most expensive hardware component in utility-scale PV projects. In lock-step with these panel cost-reductions have been less dramatic, but still important, reductions in other costs in these projects (see the first chart in section 1 above), including inverters, mounting hardware, engineering and assembly processes.
In the following subsections we will examine the architecture and major components in a typical large PV Project.
A typical medium-to-large PV project's basic configuration can be seen below.
The DC voltage determines the length of the strings (number of panels in each string). Each panel must be rated to withstand the string voltage, and the inverter must be specified with the string voltage as a DC input.
Until a few years ago 600 Vdc was the standard string voltage for large projects in the U.S., while Europe used 1,000 Vdc. In 2013 many U.S. authorities started allowing 1,000 Vdc string voltages, and UL standards quickly evolved to allow this. Existing IEC standards at that time allowed Europe to use 1,500 Vdc string voltage, but few projects in Europe used this voltage.
In the U.S., First Solar was the first manufacturer to develop panels rated for 1,500 Vdc, and in 2014 deployed the first projects to use this string voltage. Adoption rate of this voltage continues to be slow today. Although all of the required components are available, they are expensive. Also the UL 1703, Standard for Safety for Flat-Plate Photovoltaic Modules and Panels Committee only recently published requirements for the evaluation and certification of 1,500 Vdc PV modules.
There currently three panel technologies used for virtually all residential-, commercial- and utility-scale projects, as described in the table below:
Amorphous Thin Film
16% to 22.5%
14% to 16%
6% to 8%
9% to 13%
High-Temperature Performance Loss
25 to 30 years
10 to 20 years
10 to 15 years
Global Market Share
*Others include cadmium telluride (CdTe) and copper indium gallium diselenide (CIGS).
The main criteria for installations that are not space-limited is the amount of total output of an array per dollar project installation plus maintenance costs. Even though they require a larger array, polycrystalline projects generally win this battle (over monocrystalline), although it is still close.
Inverters are generally the second largest capital expenditure for large PV projects (after the panels). There are two architectures used for inverters in large projects. The block diagram in section 2.1 uses a central inverter, and these have power ratings ranging up to 4 MW. The alternative is to use a smaller "string inverter" for each string, and then feed the AC power from all strings into the MV transformer. The architecture that is best for a given project will be mostly dictated by economics.
Manufacturers of large solar inverters include GE, ABB, Schneider and SMA, and most of these manufacturers make both string and central inverters. Efficiency for these inverters ranges from 97% to 99%.
2.4.Limitations and Potential Solutions
There is really only one major limitation for PV generation, intermittency. In order for any electric energy source to be usable it must have a predictable and dispatchable output. There are two types of variability (1) the variability that occurs every day with the angle of the sun, and variability that occurs due to clouds temporarily blocking the sun. Dealing with the latter is more challenging, and this subsection discusses that. Note that PVs (alone) are not significantly dispatchable.
2.4.1.PV Plus Storage
Photovoltaic generation plus electric storage systems (mainly batteries) are the most effective method for dealing with solar intermittency and the inability to dispatch PV. There are several architectures for these systems ranging from totally independent subsystems to various degrees of integration. See the block diagrams below.
Schematic of independent PV plus storage system
Schematic of AC-coupled PV plus storage system
DC-coupled (flexible charging)
DC tightly coupled (PV-only charging)
The table below also came from the above-referenced NREL document.
Type of Coupling
Change in Value (Relative to Independent System)
Change in Cost (Relative to Independent)
Potentially lower value because it cannot be sited in regions with higher congestion-related prices
Higher losses when storing grid energy (due to additional transmission losses)a
Reduction in BOS costsb
DC (flexible charging)
Storage operation constrained by shared inverter
Potentially higher losses when storing grid energyc
Can store clipped solar that occurs due to ILR>1
Lower losses when storing solard
Limited to inverter capacity
Reduction in BOS costs
Reduction in power electronics costs due to shared inverter
DC—tight (PV only charging)
Same as DC-coupled flexible charging +
Storage cannot charge from low-cost grid energy
Same as DC-coupled flexible charging +
Cannot charge with grid energy to ensure full capacity value
Same as DC-coupled flexible charging +
Small (if any) reduction in battery management system cost
a Assuming independent storage is sited closer to load and incurs lower loss rates; this change is not considered in this study; the case also assumes the transmission capacity equals the sum of the PV and inverter ratings.
b Includes interconnection, permitting, overhead, engineering, labor, and land costs
c Due to remote location when compared to storage sited in a load center (Nourai et al. 2008)
d Not considered in this study
The results of the study performed by the referenced NREL document (benefits to cost ratio) are shown below for the Year 2020. This study shows that the ratio is higher for many configurations of PV plus storage as compared to PV only. Full details of this study are given in the referenced NREL document4. The blue and red percentages show the effects of PV market penetration as a percentage of overall generation.
B/C ratio for PV plus storage plants in a 2020 scenario with two different levels of PV penetration and the (a) 30% Investment Tax Credit and (b) zero ITC
2.4.2.Solar Sky Imaging
PV is not only variable, but rapidly variable. The output of a PV array changes directly with irradiance, so when a large cloud passes between the array and the sun, the output can decrease (and then later increase again) as much as 50% in seconds to minutes. When this happens some other resource needs to firm up the PV, that is, provide electric energy to replace the PV’s reduced output.
Because of the rapid variability conventional wisdom says that you need an electricity source that can respond instantly, which is generally limited to electro-chemical (batteries), electro-kinetic (flywheel), or electro static (super- or ultra-capacitors) storage. This assumes you cannot predict the cloud attenuating the sun (and thus the PV decrease) 30 seconds to a minute before it happens.
Solar sky imaging technology (a.k.a. solar irradiance microforecasting) is being developed by multiple groups. This technology is currently able to predict a cloud-shadow crossing the array 30 seconds prior to the crossing with better than 90% accuracy. With the above forecasting, fast-responding spinning reserve from internal combustion or turbine gen sets could be ramped up and ramped down to compensate for the PV variability.
Solar sky imaging technology can also allow smaller battery energy storage systems to compensate for PV variability, than would otherwise be required.
 Shayle Kann, Greentech Media, Keynote Address from GTM 2017 Solar Summit.
 Woodhouse, Michael, Rebecca Jones-Albertus, David Feldman, Ran Fu, Kelsey Horowitz, Donald Chung, Dirk Jordan, and Sarah Kurtz, 2016, On the Path to SunShot: The Role of Advancements in Solar Photovoltaic Efficiency, Reliability, and Costs, Golden, CO: National Renewable Energy Laboratory. NREL/TP-6A20-65872. http://www.nrel.gov/docs/fy16osti/65872.pdf.
No discussions yet. Start a discussion below.