Testing Photovoltaic Pavers for Roadway Applications

Concrete and asphalt are the primary materials used to construct roadways for motor vehicles, bike paths for pedestrians and bicyclists, and runways for aircraft. Solar Roadways®, Inc. (SR) in Sandpoint, ID, proposed using robust, Solar Road Panels (SRPs) as an alternative roadway material due to the potential for creating a modular, multi-functional infrastructure product with cost-savings, user-safety, power-generation, and a better alternative in terms of environmental sustainability when compared to contemporary pavement materials. Typical roadway construction materials, on average, need to be replaced every 10-15 years while also requiring regular annual maintenance to maintain proper safety standards. SR’s novel roadway material is intended to extend roadway replacement timelines, lower annual maintenance costs, and provide energy to the power grid. In this study, we tested the mechanical properties of the “SR3” model prototype SRP and evaluated its suitability as a replacement roadway material with the added benefit of generating electric power. Specifically, we tested this unique pavement material in submerged water environments, under extreme temperature conditions, and under dynamic loading conditions.


Introduction
In recent years there has been a global demand to use sustainable technologies which not only meet current power generation needs but also ensures that future power generation demands are met (Uzaro wski & Moore, 2008). The phenomena of the greenhouse effect or global warming is a h ighly debated topic today on domestic and international p latforms. These debates increase awareness about 86 http://www.scholink.org/ojs/index.php/se Sustainability in Environment Vol. 4, No. 2, 2019 the need for sustainability in the environ ment. People, industry, and governments are increasingly giving consideration to the environ mental effects of the materials used in paving roadways. Thousands of miles of roads are paved every year using traditional materials like asphalt or steel-reinforced concrete (Alkins et al., 2008;Bocci et al., 2011;Bowers et al., 2015;Diefenderfer et al. , 2015;Gerbrandt et al., 2000;Gu et al., 2015;Gu et al., 2019;Grilli et al., 2012;Horvath & Hendrickson, 1998;Miliutenko et al., 2013;Stimilli et al., 2013;Thenoux et al., 2007;Po lston, 2004).
In addition, the importance of energy security cannot be u nderstated in today's global economy and need for increased national security. Decentralized power production and micro grids are far mo re secure and sustainable methods of power production, as compared to any centralized power source, even centralized solar arrays. These traditional photovoltaic based renewable energy systems require that large swaths of land be dedicated to power generation. SR's Solar Road Panels shown in Figure 1, will enable photovoltaic-based power production in readily availab le areas , wh ich have already been removed fro m wilderness (i.e., roadways, sidewalks, parking areas, driveways , etc.) where vast amounts of available surface area outweigh the power production efficiency of any single panel.

Manufacturing Process-"SR3" Prototype
The "SR3" prototype Solar Road panels were manufactured by SR using a proprietary process consisting of hexagonal tempered glass plates, photo Vo ltaic (PV) cells mounted on a custom circu it board, and a novel poly mer adhesive layer. The entire structure was laminated together, with the top surface being fused with a novel poly mer adhesive layer, resulting in an engineered rough surface designed to increase the glass surface's coefficient of friction to prevent slipping. In addition to the solar power capability, the "SR3" prototypes contain Light Emitting Diodes (LED), and an internal heater circuit. The LEDs and heaters are intended to eliminate the need for painting roadways and reduce snow removal efforts in colder climates. An image of three of the SRP prototypes is shown in Figure 1. An insulated electrical cab le was imbedded in each paver and used to electrically connect each panel to the control circuit wh ich enabled monitoring and control (Brusaw, 2016).

Moisture Conditioning Duration
The mo isture conditioning duration test was conducted to investigate the mechanical properties and operation of the SRPs under wet conditions. The SRP glass surfaces represent impermeab le surfaces so that moisture has a limited effect. Therefore, the moisture conditioning test was primarily to evaluate the exposed polymer layer between the impermeable glass plates. The ASTM Active Standard D570-98(2010) e1, "Standard Test Method for Water Absorption of Plastics ", was used to evaluate the effects of water absorption or humidity exposure of the poly mer layer (ASTM Standard D570-98(2010) e1, 2014). Th is test method was intended to determine "the relative rate of absorption of water by polymers when immersed and applies to all types of polymers, including cast, hot -mo lded, and cold-molded resinous products, and both homogeneous and laminated polymers in rod and tube form and in sheets 0.13 mm (0.005 in.) or greater in thic kness". The test has two primary objectives: first, to identify the proportion of water absorbed by the polymer material and second, to identify the effects of exposure to water or humid conditions on the properties of the polymer material such as electric al, mechanical properties and on dimension, appearance, etc.
Moisture absorption affects the water content of the poly mer and is d irectly related to electrical conductivity, mechanical strength, dimension, and physical appearance. The amount of absorption depends greatly on the type of water exposure (i.e., immersion or exposure to high humid ity), the shape, and the properties of the polymer. The moisture testing was conducted on full-size SRPs not small test samples as described in the ASTM Active Standard.
The moisture conditioning testing was accomp lished using a single 300 -gallon alu minu m tank where all three pavers were simu ltaneously submerged with enough water to cover the pavers completely with approximately one-inch of water above the paver's surface. This ensured that a constant water pressure 88 http://www.scholink.org/ojs/index.php/se Sustainability in Environment Vol. 4, No. 2, 2019 was exerted on each of the pavers throughout testing. Figure 2 illustrates paver position and spacing in the tank. Also shown in Figure 3 is the wire and connector placement above the waterline during testing.  then subjected to long duration mo isture testing as follo ws: 1) a 24 -hour test, 2) a seven-day period, 3) a 14-day period, and 4) additional 14-day periods as needed. The pavers were submerged for the required time period(s), removed fro m the water tank, dried with a lint -free towel, weighed, and then moved for electrical testing (i.e., LED functionality). The apparatus for weighing the pavers was constructed fro m wood and the following steps were fo llo wed to mit igate any negative effects: 1) the fixture was stored in the temperature-controlled EMSTL at Marquette the entire t ime, 2) for every measurement the pavers were dried with a lint-free cloth prior to being placed on the test jig, and 3) the 89 http://www.scholink.org/ojs/index.php/se Sustainability in Environment Vol. 4, No. 2, 2019 scale was tarred with the fixture weight prior to the test pavers being placed on the jig and weighed.
Once electrical testing was comp lete, the pavers were again weighed and placed back into the water tanks to begin the next test cycle. Moisture testing was accomplished until the measured weight increase, after three consecutive weigh-ins, was less than 0.1% of the total paver weight or 5 grams, whichever was greater. For this test, six SRPs were electrically tested/baselined (i.e. , LED functionality) and then placed in two separate 300-gallon tubs of water inside an ESPEC walk-in environ mental chamber. The first tub contained three SRPs in fresh water, to assess mo isture conditioning capabilities, and the second tub contained three SRPs in a saltwater solution (i.e., ~ 3 % by weight NaCl) to assess performance in an expected real-world corrosive environ ment. The test consisted of 10 cycles where the ESPEC chamber temperature was set to 20 °C for 48 hours and then increased to 50 °C for 48 hours. This appro ximate five-day period constituted one test cycle. At the end of each test cycle, the SRPs were removed fro m the water tanks, dried with a lint-free towel, weighed, inspected for breaches in the physical structure, and then electrically tested. The apparatus for weighing the pavers was constructed fro m wood and the following steps were fo llo wed to mit igate any negative effects: 1) the fixture was stored in the temperature-controlled Engineering Materials and Structural Testing Laboratory (EM STL) at

Freeze/Thaw Cycling
Marquette University (MU), 2) fo r every measurement the pavers were dried with a lint-free cloth prior to being placed on the test jig, and 3) the scale was tared with the fixture weight prior to the pavers being placed on the jig and weighed. Once electrical testing was co mplete, the pavers were placed back into the water tanks to begin the next test cycle. These cycles were repeated for 10 iterations. environmental chamber shown in Figure 4.

Heavy Vehicle Simulation (HVS)
The Heavy Vehicle Simu lation (HVS) test was conducted to investigate the mechan ical properties and evaluate LED operation of the SRPs. The testing was performed in accordance with recognized industry standard practices for paved surfaces. Testing was performed on the MU campus in the SR Pilot Project area located on the south side of Engineering Hall as shown in Figure 5. Limited electrical testing was accomplished to verify paver operation before and during HVS testing and consisted of verify ing LED operat ion.

Figure 5. Heavy Vehicle Simulati on (HVS) Test Facility Located at Marquette Uni versity
Heavy vehicle traffic is a contributing factor to roadway material failures. As mentioned before the ESALs on concrete pavements are inversely related to the static modulus of sub grade reaction k-value.
The plate-load test was originally used for determin ing static subgrade k-values; however, present analysis methods commonly back-calculate subgrade k-values using center-slab surface deflections produced by a Falling Weight Deflectometer (FWD), which produces deflections that simulate wheel loads moving at highways speeds. The FWD deflection-based back-calculated subgrade k-value is typically considered as the dynamic k-value, wh ich is commonly assumed to be appro ximately twice that of the static k-value. When used in the context of the AASHTO concrete pavement design, converting from wheel loads traveling at highway speeds to creep speeds may be simulated by a 50% reduction in the design subgrade k-value. Co mparing allowable ESA ls over a range of design subgrade k-values yields a speed related damage factor of appro ximately 1.25, i.e., the slowly moving wheel loads are approximately 1.25 times more damag ing that loads traveling at highwa y speeds. Comb ining the noted load and speed effects, each pass of the HVS can be equated to approximately 1.55 ESA Ls (i.e., 1.24*1.25 = 1.551).

Moisture Conditioning Duration
The testing was conducted between 17 Ju ly 2017 and 29 Augu st 2017 and the mo isture conditioning duration testing resulted in essentially no measurable weight gain using a 100lb load cell with 0.02lb resolution. In one instance (S/N 3A), the post-submergence weight increased by 0.02lb after the one-week test period, but the paver's weight returned to its original baseline value after the next submergence period. Based on this result, a second two-week test period was acco mplished resulting in a consistent 0.04lb weight increase in each of the three test articles.
In all but one case, the LEDs were operational pre and post testing for each moisture duration test period. After the second two-week submergence test, paver S/N 3A failed electrical testing. Since the failure could not be attributed direct ly to increase mo isture content, SR conducted an in-depth 92 http://www.scholink.org/ojs/index.php/se Vol. 4, No. 2, 2019 investigation (including photovoltaic testing using the apparatus shown in Figure 8) and determined that a corroded wire (Figure 6) was the root cause of the observed failure. The wire was exposed, due to a wire shielding flaw during manufacturing ( Figure 7) and acted like a wick during the mo isture duration tests. Improved manufacturing processes and use of more robust components will p revent similar future failures. The wires were cleaned/repaired and then LED functionality and photovoltaic tests were conducted on all three panels at the SR facility. Figure 8 shows the underside of the SR solar tester. It contains seven 500W lamps for a total of 3500W. The entire apparatus was placed on top of a single Solar Roadway panel during testing and the results are provided in Table 1.

Freeze/Thaw Cycling
Freeze/Thaw testing was conducted between 23 February 2018 and 20 April 2018 on six "SR3" solar pavers with S/N's 1A-1F. For all specimens, there was no gain in paver mass detectable to the nearest 0.02 lb. In addit ion, no physical defects were noted after the 10 freeze/thaw cycles.
Finally, in all cases, the LEDs were operational in pre and post testing of each of the Freeze/Thaw test cycles. The six SRPs tested showed no sign of any adverse effect after rigorous Freeze/Thaw testing.
The pavers neither gained any mass due to water infiltrat ion nor showed any sign of physical distress after the test regimen.

Heavy Vehicle Simulation (HVS)
HVS testing on six SRPs with S/ N's 015A-015F began on 9 July 2018 and ended on September 30 approximately two hours of paver power or color reset. Appro ximately one-third of the LEDs in pavers 015D and 015F remained operational. LED status is depicted on Figure 9. The observed LED performance variations were again attributed to variations in the SRP p rototype's manufacturing process, previously shown in Figure 5 (i.e. , cracked internal PV cells, internal in-poly mer "bubble" formations, variances in the external "etched " surface grip/traction features, etc.), and wire corrosion due to ingress fro m cracked wire shields during continuous outdoor operation.

Conclusions
Three different tests were conducted to evaluate the mechanical properties of SRPs. The results show the current SRPs to be robust, resilient, and functional when subject to "real-world" test conditions.
None of the observed anomalies were catastrophic failures and all were related to variations due to the manufacturing process. Finally, no direct co mparis on to concrete, asphalt, or other plastic materials can be made due to the unique nature of the SRPs and the materials used in their manufacture.