A CASE STUDY OF A RESIDENTIAL PHOTOVOLTAIC SYSTEM WITH MICROINVERTERS Clifford K. Ho Sandia National Laboratories P.O. Box 5800 Albuquerque, NM 87185 E-mail: ckho@sandia.gov ABSTRACT This paper presents a case study of a 3 kw photovoltaic system with Enphase microinverters. A brief introduction to microinverters and central inverters is first presented, followed by an overview of the installation, features, operation, monitoring, and costs of the system. The monthly and annual energy production during the first three years is presented and compared against modeled results using the System Advisor Model. Various factors that have impacted the performance (e.g., shading, weather, temperature, wind, tilt, orientation) are discussed, and the performance of the system is compared against a simulated optimized system. The optimal sizing of microinverters is also discussed to explain why the power rating of the microinverter may be less than the power rating of the module to achieve maximum energy production. 1. INTRODUCTION Solar inverters convert direct current (DC) electricity generated from photovoltaic (PV) modules to alternating current (AC) electricity that can be used by devices and appliances. A conventional central inverter is connected to a string of multiple PV modules, whereas a microinverter is connected to a single PV module. Some pros and cons associated with microinverters relative to central inverters that have been marketed and discussed include the following (1),(2): Pros More energy produced with partial shading Greater reliability (e.g., 25 year vs. 5 year warranty) Ease of installation Safety (no high-voltage DC lines) Monitoring of individual modules Cons Higher capital cost Placement is difficult to access (behind module) This paper presents a 3-year case study of the first commercial PV system installed in New Mexico employing Enphase microinverters (see Fig. 1). The installation, operation, and features of the system are first discussed. The energy production of this system during the first three years is then presented and compared to model predictions using the System Advisor Model (3). Optimized designs of the modules and microinverters are then presented, followed by a discussion of the costs and return on investment. Fig. 1: Enphase microinverter (from Enphase.com). 1
2. INSTALLATION, OPERATION, AND FEATURES 1.1. Materials The 3 kw grid-tied PV system consists of the following components and features: 15 PV modules o 200 W Sanyo HIP- 200BA3 o Each module oriented 22 degrees west of true south o Tilt ~27 deg (top array), ~30 deg (bottom array) Enphase Microinverter o 200 W (M200-32-240) o 15 year warranty Fig. 3: 200 W Sanyo HIP-200BA3 modules mounted on racks over each microinverter. 1.2. Installation The 3 kw PV system was installed by Sunergy (now CleanSwitch) and commissioned with the Public Service Company of New Mexico (PNM) in December 2008. Ten 200-W modules and microinverters were installed on the second-floor roof, and five 200-W modules and microinverters were installed on the first-floor roof (Fig. 2- Fig. 4). The Spanish tiles had to first be removed so that the racks could be mounted to the roof trusses. The microinverters were then mounted to the racks at locations corresponding to the position of each module (Fig. 2). The modules were then mounted to the racks and connected to each microinverter. Note that a chimney to the east of the modules shades several of the modules during the morning hours at different times during the year (Fig. 3-Fig. 4). The microinverters are connected to a disconnect switch, a renewable energy credit (REC) meter, and a net meter on the south side of the house to monitor energy production. Fig. 4: Completed 3 kw PV system with 15 modules and microinverters (10 on the 2 nd -floor roof and 5 on the 1 st -floor roof). 1.3. Operation and Monitoring The operation of the 3 kw PV system with microinverters is summarized in Fig. 5. AC power produced by each module is monitored via AC wiring in the house and transmitted through the internet to an Enphase Enlighted website. Fig. 2: Enphase M200-32-240 microinverters (one per module) mounted to racks. 2
Fig. 5: Operation and monitoring of system performance (from Enphase.com). Fig. 6 shows an image of the entire 3 kw system, with 10 modules on the second-floor roof and five modules on the first floor roof. The current power production and energy production produced over various intervals (day, week, month, lifetime, or user-specified) are also displayed. The power production of each module is displayed on each module (Fig. 7), and energy produced, DC current, DC voltage, and temperature can also be displayed for each module. The luminance of each module also conforms to the power being produced (modules producing more watts appear brighter). Fig. 7: Individual module performance can be monitored including power, energy produced, DC current, DC voltage, and temperature. Fig. 8 shows the system web page when several of the modules were shaded in the morning. The modules that were shaded show a power production of less than 20 W (from diffuse lighting) while the modules exposed to sunlight show a power production of 116 120 W. This illustrates the ability for modules to produce power independently of each other when using microinverters connected to each module. With central inverters, if a single module is shaded, all of the modules connected on that string would be negatively impacted. Fig. 6: Website to monitor system performance. 3
Fig. 8: Shading of individual modules does not impact the performance of other modules. Automated alerts are also sent by e-mail if anomalous conditions are recorded. After a PNM power outage, an automated e-mail alert was sent to me with the title AC Voltage Out Of Range. The message stated that the microinverter detected that the AC voltage coming from the utility was either too low or too high as specified by applicable regional standards and that the microinverter would remain offline until the utility had been within acceptable limits for a period of time. The condition was self-correcting. I received another e-mail alert in which I had to call Enphase to report the error, and technical support was able to reset the microinverter remotely and get it working again. 3. ENERGY PRODUCTION The energy production of the 3 kw PV system was monitored and recorded from January 2009 to December 2011. Comparisons between the energy production reported on the Enphase website and the energy production recorded from our REC meter revealed that the Enphase-reported energy production was consistently ~4% higher than that recorded from the REC meter. I spoke with a technical support representative from Enphase on the phone and he stated that the accuracy of the monitoring system was ±5% while the REC meter was ±2%. He also stated that it was common for the Enphase monitoring system to report on the high side. Fig. 9 and Fig. 10 show the monthly and annual energy produced from the system in 2009, 2010, and 2011. Predictions from the System Advisor Model (SAM) (3), which used built-in weather and solar irradiance data from a typical meteorological year (TMY2) in Albuquerque, NM (Latitude = 35, Longitude = -106.6, Elevation = 1619 m), are also shown for the 3 kw system. Results show that while there is considerable monthly variability in the energy production, the annual energy production was consistent with an average of 5,270 kwh and a standard deviation of 65 kwh. The SAM predictions yielded an annual energy production of 5,230 kwh, just 0.93% above the measured average annual energy production. It should be noted that the SAM model did not include partial shading impacts from the chimney. Fig. 9 shows that there is a trend in the measured monthly energy production during the first three years. The greatest energy production is in the spring (April May) and late summer (August September). During these times, the maximum solar irradiance is received by the modules based on their tilt and orientation. In addition, the temperature is cooler, which increases the power production of the modules. The wind is also stronger in April/May in Albuquerque, which also helps to keep the modules cooler. The SAM models predicts similar trends. Monthly Energy Produced (kwh) 600 500 400 300 200 100 0 SAM Model Data - 2009 Data - 2010 Data - 2011 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Fig. 9: Monthly energy production during the first three years of the 3 kw PV system compared to predicted energy production using SAM. 4
Annual Energy Produced (kwh) 6000 5000 4000 3000 2000 1000 In addition to the module tilt and orientation, the size of the microinverter was also investigated. Fig. 12 shows that the inverter efficiency decreases as the power output decreases relative to its rated (nameplate) rating. Therefore, it is desirable to have the inverter operating near its rated output power. 0 SAM Model 2009 2010 2011 Fig. 10: Annual energy production and SAM predictions during the first three years. 4. OPTIMIZATION The orientation and tilt of the 3 kw system was varied in SAM to determine an optimal configuration using the TMY2 weather file for Albuquerque, NM. Results of the parameterization showed that a tilt of 35 and an orientation of 10 east of true south yielded the greatest annual energy production of 5,390 kwh (Fig. 11). However, this is only ~3% greater than that predicted using the existing configuration. A reason for the optimal orientation being faced slightly to the east may be because of the significant cloud cover in the afternoons during the monsoon season (July through August). True South Existing System (22 deg west of true south, <30 deg tilt) Optimized System (10 deg east of true south, 35 deg tilt) Fig. 11: Illustration of existing and optimized module configuration for Albuquerque, NM, using SAM predictions. Fig. 12: Plot of inverter efficiency as a function of percent rated output power (3). While it may seem logical to combine a 200-W microinverter with a 200-W module, the overall efficiency of the microinverter may not be optimized since the power production in the morning, evenings, and during cloud cover can be significantly lower than the rated output power. Thus, the efficiency of a 200-W microinverter connected to a 200-W module may be lower than that of a, say, 175-W microinverter. Fig. 13 illustrates this concept by showing the hypothetical power produced from a 200-W module using either a 175-W (red plot) or 200-W (black plot) microinverter. During the mornings and evenings when the solar irradiance is less, the efficiency of the 175-W microinverter is greater, yielding greater power output. However, when the solar irradiance is high around solar noon, the power produced may be greater than 175 W, and power production will be clipped by the 175-W microinverter. Since the total energy production is equal to the area under the power production curve, if the additional area gained by the 175-W microinverter in the mornings and evenings is greater than the area lost due to clipping relative to the area generated by the 200-W microinverter, the total energy produced will be greater. Fig. 14 shows a comparison of simulated monthly energy production from the 3 kw system using different sized microinverters in SAM. Although similar, the use of the 175-W microinverters actually produced slightly more 5
energy (5,232 kwh vs. 5,225 kwh) over the course of a year than the 200-W microinverter when paired with the 200-W modules for the 3 kw system. Fig. 13: Illustration of hypothetical power produced from a 200-W module using a 175-W and 200-W rated microinverter over the course of a sunny day. Energy Produced (kwh) Power Produced 600 500 400 300 200 100 0 Fig. 14: Comparison of monthly energy production from the 3 kw system using different sized microinverters. 5. COSTS morning noon afternoon 175 W Time M200 Enphase microinverter with 200 W module M175 Enphase microinverter with 200 W module SAM Model - M200 Enphase Microinverter SAM Model - M175 Enphase Microinverter Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec The cost of inverters depends on size, technology, brand, etc. One of the disadvantages of microinverters is that they are currently more expensive than central inverters for given system requirements. Costs of traditional central inverters range have been reported to range between $0.35 - $0.50 per watt, while the costs of microinverters have been reported to range between $0.52 - $0.90 per watt (2),(4). The total cost of our 3 kw system (including installation) in 2008 was ~$23,000 after a 30% tax credit (~$33,000 before the tax credit). PNM pays me 13 cents for every kwh generated by my PV system (to receive credit towards the Renewable Energy Act requirement of generating 20% of their electricity from renewable sources by 2020). This amounts to a payment of nearly $700 per year (using the average annual energy produced from my system for the first three years). In addition, because our household uses about 5,000 kwh of electricity per year, we save about 5,000 kwh x $0.08/kWh = $400/year. In total, our savings/revenue is about $1,100 year. Therefore, it will take roughly $23,000 $1,100/year = 21 years to recover the cost of our system. It should be noted, however, that the total installed cost of my system before tax credits was over ~$10/W. Colleagues who have recently installed similar systems paid just ~$5/W - $6/W (before tax credits), nearly half the cost of my system installed just three years ago. In addition, tax credits in New Mexico are now at 40% (30% federal and 10% state). This increased in 2009 from a 30% combined tax credit, just days after my system was commissioned on December 19, 2008! 6. CONCLUSIONS A case study of the first commercial PV system employing Enphase microinverters in New Mexico has been presented. The installation, features, operation, performance, and costs have been discussed. During the first three years from January 2009 to December 2011, the 3 kw system produced an average of ~5,270 kwh per year. Predictions using the System Advisor Model, which used the same system configuration and a typical meteorological year (TMY2) weather file for Albuquerque, NM, yielded very similar results. Monitoring revealed that shading of individual modules from a nearby chimney did not impact the power produced from adjacent modules, illustrating one of the benefits touted for microinverters. A parametric analysis using SAM revealed that an optimized system configuration with a different module tilt and orientation yielded an increase in annual energy production of ~3% relative to the existing system. A 6
discussion of optimal microinverter sizing was also presented that illustrated the efficiency dependence of inverters on output power. Finally, the costs of microinverters vs. central inverters was briefly discussed, followed by an economic assessment of the 3 kw PV system. ACKNOWLEDGEMENTS Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy s National Nuclear Security Administration under contract DE-AC04-94AL85000. This manuscript has been authored by Sandia Corporation under Contract No. DE-AC04-94AL85000 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. REFERENCES (1) Enphase Brochure, 2011, www.enphase.com (2) Solar micro-inverter, 2011, Wikipedia website: http://en.wikipedia.org/wiki/solar_micro-inverter (3) System Advisor Model, National Renewable Energy Laboratory, Golden, CO (https://sam.nrel.gov/) (4) Solar Central Grid Tie Inverters vs. Microinverters Advantages and Disadvantages, Green World Investor, April 14, 2011, http://www.greenworldinvestor.com/2011/04/14/solarcentral-grid-tie-inverters-vs-microinverters-advantagesand-disadvantages/ 7