Case study: 100kW Floating PV solar system

PV System description

This is a 100kW PV System off-grid without battery operating on a floating platform some short distance from Singapore's shores.  It has 4 zones each 4 zones each with its own set of inverters to run the pumps, lighting etc. Figure 1 below illustrates the 4 sub-system areas with colored outlines in Red, Green, Yellow and Magenta. With 16 strings connected to 4 inverters, the entire testing takes about 1 hour to complete considering the initial walkabout to familiarize the system layout and time to decide and select the cable areas to measure. Subsequent checks will likely take a fraction of the time.

The PV system uses two module types: 310W modules at zones 1 & 4 and 330W modules at zones 2 & 3. The panels act as roof for the single level operation. As shown in photo, zones 1 & 2 are oriented with a tilt in one direction while zones 3 & 4 are tilted in the reverse direction. There is no ready access to actual electrical schematics but each zone is observed as having 4 strings or 16 strings in total; each string is comprised of 22 modules in series.

Test Approach

The normal way is to fix the sensor co-planar to the string to be tested and head off to the inverter to measure the DC incoming cables. However, the given space to access the inverters is very tight and a decision is reached to test each string with a ladder.

Using an improvised quick lock and release holder, sensor powered by a 5V power-bank is mounted with similar inclination to the string to test. Next, each string is then measured with a DC clamp ampere meter.

When reading on clamp meter stabilizes without fluctuation, the value is entered into the PVBuddy mobile app and with a final click of the ‘CHECK’ button, result is ascertained in a second or two. Data is then logged for an accepted check.

The sensor has long range wireless connectivity to app and as long as all strings are in same inclination, user need only to focus on clamping each string and check the results.

Figure 2. Sensor with powerbank placed co-planar to string under test

The intensity of the sun or irradiance ranges from 256 W/m2 to 1059 W/m2 when test results were conducted; the assessment results of same string under differing irradiance remains consistent. More significantly was the time taken to conduct and complete the tests: all 16 strings including retests to confirm took only 30mins or less.

Findings, Trouble-shoot and Optimization

Most of the strings registered OVER the expected computed in range of 2% to 12%. This is not uncommon and comforting to know the panels in each of these strings are operating beyond the nameplate expected ratings. Why? One possible reason is module makers tend to sell modules slightly over or above the nameplate label rating to ensure they will meet the minimum contractual performance and account for short term solar cell degradation such as Light Induced Degradation.

One set of strings in Zone 1 however, exhibited gross under-performance with recorded degradation range from 38% to 44% as shown in figure 4. A recommended problem solving approach is to visually determine if the panels are not dirty or shaded by nearby obstructions; then ascertain any alarms or review event logs at inverter. Since the strings are all connected to same inverter and are having similar under-performance, the inverter is likely having some issues. We were not told of any technical faults for this system beforehand and this suspicion was later confirmed by foreman. We used PVBuddy to quantify the degree of under performance of these strings.

Figure 4. Actual ampere reading vs Expected and % degradation at Zone 1 strings

Cost impact

Let’s crunch some numbers for these 4 strings to this particular inverter.

Each string has 22 nos of 310W modules in series with theoretical Wp of 6820W.

Singapore enjoys average Peak Sun Hours (PSH) of 4.38

Present electricity kWh rate is $0.287

Figure 5. Table of cost impact due to power loss

As shown in the table above, the daily, monthly and annual loss amounts to $14.15, $424.37 and a whopping $5,092.45 respectively.

The urgency to attend to the loss will help curb the loss which in this case is to repair or replace the defective inverter.

Conclusion

The objective of the above exercise is to assess how well each string is performing under load and ultimately consider the necessary steps to optimize the PV system power generation further. Given that we have no prior knowledge of the said pv system, PVBuddy test kit performed very well and its ability to quantify the extent of under, within or over performance will help give insights to the PV system behavior.

The speed to check, rank and data log each string impressed the owner and more importantly, the findings will allow user to conduct troubleshooting to isolate any areas to optimize and make informed decisions as to how to maximize their returns and investment in their solar assets.

For more information on PVBuddy test kit, please visit us here www.pvbuddy.com