Predicting location and evaluating progression of clogging in a permeable pavement parking lot

Abstract

In 2009, a permeable pavement research and demonstration site was constructed at the Edison Environmental Center, Edison, NJ. Infiltration testing of three original permeable parking rows through August 2012 indicated that clogging occurred along the upgradient edge of these pavements from runoff that drained from adjacent impermeable driving lanes.

A subsequent infiltration testing data collection effort from April 2017 through March 2020 focused on permeable interlocking concrete pavers (PICP) that replaces one of the original permeable surfaces. While the original infiltration study through 2012 used random locations throughout the permeable parking rows, the newer study targeted upgradient edge to identify where clogging would occur. Testing locations along the upgradient edge were selected based on a high-resolution survey (HRS) of the parking lot performed in December 2014. The HRS identified three low spots along the upgradient edge that eventually clogged in the new PICP infiltration study.

The HRS may assist with maintenance routines. The newer study also supports the conclusion of the earlier study with regards to truncating the infiltration testing method, particularly for maintenance assessments.

Introduction

The Environmental Protection Agency’s (EPA’s) Office of Research and Development (ORD) has monitored a permeable pavement research and demonstration parking lot at the Edison Environmental Center (EEC), NJ since December 2009, one month after first use as a parking lot for EEC employees and visitors. This site originally had three different permeable pavements including permeable interlocking concrete pavers (PICP), pervious concrete and porous asphalt. The permeable pavements were limited to parking spaces while the adjacent driving lanes, which were impermeable asphalt, drained to the permeable surfaces as the entire 0.4 ha parking lot was designed with a 1.6% slope. The nominal design ratio of watershed area to permeable area is. 0.66 for each permeable surface.

The infiltration testing method used for all three pavements was based on ASTM C1701 (ASTM, 2009) with some amendments described in the Methods section; at the time, ASTM C1701 was the only testing method available for permeable pavement (ASTM C1781 was approved in 2013). The average infiltration testing rates from December 2009 through April 2010 were 40,000 mm/hr for pervious concrete, 24,000 mm/hr for PICP and 2,000 mm/hr for porous asphalt (Borst et al. 2010) and at the time were similar to literature values for newly installed permeable pavements, i.e., 40,000 mm/hr for porous concrete and 20,000 mm/hr for PICP (Bean et al. 2007) and 4,300 mm/hr for porous asphalt (Ferguson, 2005).

Infiltration testing continued through August 2012 and declining infiltration rates for the three surfaces were attributed to clogging occurring along the upgradient edge of the permeable pavements due to sediment laden runoff from the adjacent driving lanes (Brown and Borst, 2014). Figure 1 is a picture taken during a visual assessment of the extent of clogging of the PICP surface on 12/27/2012 which was presented graphically in Brown and Borst (2014).

Figure 1.

Permeable interlocking concrete pavers during visual assessment of clogging on 12 /27/2012

Clogging along the edge has been observed by others as well. Kazemi et al. (2017) describes progression form the upgradient edge. Winston et al. (2016), Chen et al. (2020) and Simpson (2021) also noted significant declines in infiltration rates along the edge of permeable pavement. Further while decline in infiltration rates was most pronounced along the upgradient edge, least-squares linear regression analyses of all infiltration tests indicated declining mean-monthly infiltration rates for each permeable surface (Brown and Borst, 2014). Chen et al. (2020) also noted declines in infiltration rates generally throughout two permeable pavement systems.

Besides observed clogging along the upgradient edge, and general trend for declining mean-monthly infiltration rates, the initial infiltration testing period, i.e. December 2009 through August 2012, indicated that the three permeable surfaces were still infiltrating and therefore the need for cleaning maintenance of the entire surface was not anticipated for decades. This observed PICP clogging progression at the EEC was slower than at other EPA monitored sites. For an EPA research project in Louisville, KY, the drainage area to the working permeable pavement width ratio (m²/m) was more than two orders of magnitude larger than the design ratio at the EEC and required cleaning maintenance after several months (Brown and Borst, 2013). At another EPA PICP research site monitored in Kansas, the contributing area to permeable area was also larger than that of the EEC, i.e., 120 m²/m compared to 8 m²/m, respectively, and cleaning maintenance of the permeable surface was required annually due to clogging (Razzaghmanesh and Borst, 2018).

Due to severe surface raveling, the porous concrete parking rows at the EEC were replaced in July 2016 with a different PICP. Limited infiltration sampling of the new PICP with smaller gaps between paving blocks indicated it had a slower infiltration rate than the original PICP. The original PICP blocks (ECO Paver, EP Henry, Woodbury, New Jersey) had dimensions of 0.155 m (6.1 in) × 0.216 m (8.5 in) with thickness of 0.0794-m (3.125-in) and had a 12.7-mm (0.5-in) wide gap between blocks creating a surface open area of approximately 11%. The newer PICP blocks (Belgard 80mm AQUAROC Natural) had the same nominal thickness but differed both in having a smaller gap width between pavers of 8 mm (0.3 in) and surface block dimensions of 0.11 m (4.5 in) x 0.23 m (9 in) yielding a smaller surface open area of approximately 9%. Both pavers were placed on a double bedding layer comprised of a top layer of 0.05 m (2 in) American Association of State Highway and Transportation Officials (AASHTO) No. 8 aggregate and underlying layer of 0.1 m (4 in) of AASHTO No. 57 aggregate that were placed over a common reservoir of AASHTO No. 2 aggregate. The AASHTO No. 8 aggregate also filled gap spaces between the original pavers while the newer pavers with a narrower gap used a smaller specified aggregate of AASHTO No. 9 to fill gaps.

The placement of the new PICP allowed for the testing of three hypotheses: 1) would targeted infiltration testing of the upgradient edge show the progression of clogging into the new PICP surface; 2) would a different PICP application with a smaller surface open area provide lower initial infiltration rates and more pronounced losses in infiltration rates over time; and 3) could locations of severe clogging along the upgradient edge be predicted in advance of any infiltration testing. Additional infiltration analysis would also allow further analysis on the Brown and Borst (2014) recommendation that the pre-infiltration testing is sufficient for most maintenance requirement determinations.

Infiltration testing started on the new PICP in April 2017 and proceeded on a quarter-year basis. The original infiltration testing effort used random sampling of the three permeable surfaces (Borst et al. 2010 and Brown and Borst, 2014). The surface infiltration testing of the new PICP targeted the upgradient edge. Additional measurements linearly downgradient of edge points were tested to track the progression of surface clogging due to runoff from the impermeable driving lane, as per recommendations in Brown and Borst (2014). For comparative purposes to the previous PICP study, the same modified version of ASTM C1701 was adopted for infiltration testing of the new PICP.

Infiltration sampling locations along the upgradient edge were chosen for the new infiltration testing effort based on a high-resolution survey (HRS) of the parking lot performed in December 2014. The HRS provided survey elevations (ft) above mean sea level (AMSL) down to 5 mm (0.02 ft) and corresponding elevation contour intervals to 0.03 m (0.1 ft). The HRS identified the drainage areas for each permeable surface that included not only the adjacent northern driving lanes, the primary contributing areas for each permeable surface (as designed), but also drainage from the western driving lane between building and parking lot, as previously described in Brown and Borst (2013 and 2014). The contour lines through the permeable pavements and northern driving lanes were generally parallel to the upgradient slope but there was evidence of microtopography. Figure 2 is a drawing of the parking lot featuring details of driving lanes, curbing and parking rows with brick hatching for study areas of concern; the cross-section A-A presents the design elevations AMSL and the results of the of HRS survey elevation contour lines are presented in plan view at 0.15 m (0.5 ft) intervals.

Figure 2.

Plan of permeable pavement parking lot with elevation contour lines and cross-sectional views

Due to the slow progression of clogging previously observed on the original PICP, it was anticipated that the HRS would assist with predicting areas of severe clogging on the new PICP. Most sampling locations were chosen where the HRS suggested concentrated flow would occur along the upgradient edge due to lower elevation, while other locations were chosen with an expectation of less chance of clogging due to higher elevation. Though contours were provided at elevation intervals 0.03 m (0.1 ft), differences between low and high spots along the upgradient edge of pavers was on the order of millimeters (hundredths of a foot). Sampling locations were also chosen away from any direct interference by excess runoff from eastern and western driving lanes and should have only been receiving runoff from the driving lane directly upgradient to sampling locations.

Methods

The surface infiltration rates were measured using a modified version of ASTM C1701 method for porous concrete (ASTM (2009) Standard Test Method for Infiltration Rate of In-place Pervious Concrete West Conshohocken, PA) as detailed in Brown and Borst (2014). Neoprene was used instead of plumber’s putty to form a seal between the permeable surface and the 300 mm (12 in) diameter cylindrical ring containing water for the infiltration tests. In addition to the piece of neoprene cut to cover the edge of the cylindrical ring, neoprene pieces were cut to insert between paver gaps to provide a seal and prevent lateral migration of water for infiltration tests on both PICP surfaces. ASTM C1701 is a two-step procedure that uses 3.6 kg of water for a pre-wet test followed by the standard infiltration test of either 18 or 3.6 kg, depending on whether pre-wet test was completed within 30 sec. Another modification was to terminate testing after 1800 sec and measure remaining water level in cylindrical ring; if this occurred during the pre-wet test, only the pre-wet test was performed. Others have truncated infiltration tests for clogging sites, i.e., Winston et al. (2016) stopped adding water after 900 sec.

Testing locations were specified as a distance from either eastern or western curb and from the north edge of the permeable pavement surface to the outer edge of the 300 mm (12 in) diameter cylindrical ring used for testing.

The initial four quarterly infiltration tests were comprised of 15 individual infiltration tests at five locations along the upgradient edge of the new PICP and two additional infiltration tests per location farther down gradient. The first down gradient tests were performed without spacing interval as the next two tests were adjacent to down gradient edge of sampling cylinder, i.e., 0.3 m and 0.6 m from upgradient edge. Upon further review of ASTM C1701 (ASTM, 2009), which indicated infiltration tests should be at least 1 m apart, remaining down gradient tests were measured 1 m from edge of pipe at upgradient position to edge of pipe at down gradient position. The remaining quarterly infiltration tests were then comprised of 16 individual infiltration tests with eight locations along the upgradient edge and one paired downgradient test per upgradient location. During the sampling period from April 2017 to March 2020, 175 infiltration tests were performed.

Infiltration rate for performance criteria, i.e., infiltration or rejuvenated infiltration rates after maintenance, or clogging or severely clogged were inconsistent in the literature without much reasoning for values used. Drake and Bradford (2013) cited 250 mm/hr as a criterion for needed maintenance by the Interlocking Concrete Pavement Institute (ICPI) but opted to perform statistics on maintenance with a break point of 50 mm/hr (rejuvenated above this mark and clogging below). Winston et al. (2016) indicated one surface as heavily clogged with a median infiltration rate of 60 mm/hr and satisfactory post-infiltration, median infiltration rate of 9,000 mm/hr. Performance criteria in this paper were based on the equation to calculate infiltration rate in ASTM C1701 with modification to cease testing at 1800sec and estimate remaining volume. The infiltration rate (ASTM, 2009) was calculated by the following equation:

$$I=KM/\left(D^{2}t\right)$$ Equation 1

where I = surface infiltration rate (mm/h); K = dimensional constant 4,583,666,000 (mm³s)/(kg/h); M = mass of water (kg); D = inside diameter of the cylinder (mm); and t = measured drain time (s). Using the above equation with two defined limits, i.e., 3.6 kg of water infiltrating over 30 or 1800 seconds, the results were grouped into three classes for analysis:

1. Infiltrating - no impact, infiltration rate larger than 5,920 mm/hr (233 in/hr)

2. Clogging - impacted, infiltration rate 5,920 mm/hr (233 in/hr) to 99 mm/hr (3.9 in/hr)

3. Clogged – severe clogging, infiltration rate 99 mm/hr (3.9 in/hr) or less

Color coding of white for “Infiltrating”, grey for “Clogging” and black for “Clogged” was used for graphical data presentation in Excel and AutoCAD (Autodesk AutoCAD LT 2015). Additional graphical and statistical analysis were performed in Excel and Statistica 9.1 (StatSoft, 2010).

When a testing location was deemed Clogged, the next test for that location was one pipe diameter or 0.3 m downgradient from the edge, with the subsequent paired downgradient test measured 1 m from the downgradient edge of the new first test location.

Findings

Quarterly infiltration testing on the new PICP began 4/27/17 (270 days after installation) and 175 infiltration tests are inclusive through 3/12/2020 (1,319 days after installation). Figure 3 shows that results of infiltration testing indicate that location along the upgradient edge is the primary factor for Clogging with all four Clogged tests occuring on the upgradient edge. Figure 4 shows that during the initial sampling event, 270 days after installation, none of the testing sites along the edge indicated Clogging, while beyond 1000 days, all sampling locations indicated Clogging. It took about 600 days after installation for the first testing locations to be deemed Clogged.

Figure 3.

Results of surface infiltration testing on new PICP

Figure 4.

Infiltration rate along upgradient edge versus days since installation

Results indicate that low lying areas were identified by the HRS along the driving lane immediately upgradient to the new PICP on the western side of the parking lot leading to concentrated surface runoff and accelerated clogging. Figure 5 shows that western side of the new PICP had more sites identified as Clogging than the eastern side and had the four Clogged infiltration tests. The HRS performed during 2014 included the damaged pervious concrete, and therefore after removal of the pervious concrete and replacement by the new PICP in 2016, the contour line nearest the edge was only predictive if portions of this contour line were in the driving lane. This is the case as shown in Figure 6 for the western side of the parking row where the 86.6 ft AMSL contour line passed into the driving lane, identifying low spots. The next upgradient contour line at 86.7 ft AMSL (0.03 m higher) was used to help predict other areas of low and high spots but was less effective at predicting Clogged sites. Additional sampling locations through March 2019 are represented in figures in the appendices.

Figure 5.

Comparison of infiltrating rates on western and eastern part of parking row

Figure 6.

Close-up plan of parking lot with three clogged testing sites and contour elevations

Figure 6 shows the three Clogged infiltration testing sites on the western side of the new PICP parking row downgradient of the impervious asphalt driving lane. Two infiltration test sites that are low points on the 86.6 ft AMSL contour line both tested as Clogged 577 days after installation. An additional infiltration test was conducted 729 days after installation on the furthest east infiltration test site (Figure 6); this additional infiltration test was also considered Clogged, hence the four Clogged tests on Figure 5 despite only three Clogged locations on Figure 6. The third Clogged site, furthest west test location in Figure 6, was tested 728 days after installation and was attributed to the low point of the 86.7 ft AMSL contour line not the 86.6 ft AMSL contour line. A subsequent infiltration test 954 days after installation slightly to the west of this location (0.13 m) was only considered Clogging, hence the overlapping orange coloring.

In comparison, no infiltration testing locations were considered Clogged for the original PICP, and only 9 of 204 measurement were considered Clogging (< 5,920 mm/hr). Eight of these Clogging locations were within or abutting an area of visual clogging from sediment accumulation identified by Brown and Borst (2014) along the upgradient edge of the original PICP, while one was 8.7 m (28 ft) from the upgradient edge but only 0.97 m (3.2 ft) from eastern curb. All other testing locations were considered Infiltrating (> 5,920 mm/hr). Despite relatively few samples defined as Clogging, a regression analysis of monthly average infiltration rate indicated a statistically significant decline in surface infiltration rate for the original PICP and the other surfaces (Brown and Borst, 2014).

As for the original PICP data, a mean monthly infiltration rate regression analysis was also performed for results of all infiltration tests of the new PICP and the results of both regression analyses are presented in Figure 7. The new PICP slope of declining infiltration rate at −316 mm/(hr-month) is steeper than the original PICP slope rate of −220 mm/(hr-month). The y-intercept of the new PICP is approximately 9,000 mm/hr lower than the original PICP confirming slower initial rates for new PICP. This result is consistent with a study of clogging dynamics by Ehsaei (2013) showing that larger gaps between pavers had larger infiltration rates and slower clogging. Drake and Bradford (2013) also indicated a paver system with smaller areas between pavers led to quicker infiltration rate declines.

Figure 7.

Regressions of monthly infiltration rates for both permeable interlocking concrete pavers

Figure 7 also shows the standard error of infiltration rate for the new PICP study is greater than the original PICP study, 1300 to 600 mm/hr, respectively, though this is due to fewer monthly tests for the new PICP study, 11 compared to 27 monthly tests for original PICP study. However, the most important result is the increase in coefficient of determination for the new PICP (R2=0.81), in comparison to original PICP (R2=0.42); this increase in the observed coefficient of determination is a direct result of the newer study focusing on the upgradient edge where clogging is predominantly taking place, compared to the original study that used randomly selected test locations.

The effects of the new targeted infiltration study compared to the original random study was confirmed by further statistical analysis. A multiple regression analysis of all surface infiltration tests (379 data points), coded by different PICP surfaces, was performed. This model of infiltration rate took the following form:

$$\text{Infiltration}\text{rate}=A+B^{*}\text{Days}+C^{*}\text{Surface}+D^{*}\text{Days}*\text{Surface}$$ Equation 2

The “Surface” coding was 1 for the new PICP and 0 for the original PICP study and the infiltration parameters were all considered statistically significant (p << 0.05). This implies the original PICP study was dependent only on time whereas the new PICP is dependent on time and surface type, which supports the result of random versus targeted study, respectively.

Besides recommending infiltration testing along the edge and linearly downgradient of these edge locations, Brown and Borst (2014) also indicated that using the pre-wet infiltration test only for PICP was acceptable for quick assessments, particularly if the objective of tesing was to reveal clogging for maintenance. When Brown and Bost (2014) compared the mean pre-wet test infiltration rate to the mean standard test infiltration rate for PICP, there was a 12% increase to the mean standard test. A similar result was found for the new PICP infiltration testing results that had an 8.6% increase to the mean standard infiltration test.

Another way to look at this is to evaluate whether the pre-wet test, which uses a 30 sec threshold to determine the mass of water for the second standard infiltration test mass, correlates with the previously defined grouping classes for the standard infiltration test of Infiltrating or Clogging, calculated as above or below 5,920 mm/hr (233 in/hr), respectively. Table 1 shows how the number of infiltration tests considered Infiltrating (> 5,920 mm/hr) or Clogging (< 5,920 mm/hr) compared to number of pre-wet tests that were under or longer than 30 seconds. For both PICP surfaces, correlation between pre-wet tests and defined grouping class exceeded 99% as only one case differed for each surface.

Table 1.

Comparison of infiltration test results to pre-infiltration tests

Condition	Number of Original PICP Infiltration Test	Number of New PICP Infiltration Test
Infiltrating (>5,920 mm/hr (233 in/hr))	195	90
Pre-wet test (<30 Sec)	196	89
Clogging (<5,920 mm/hr (233 in/hr))	9	81
Pre-wet test (>30 Sec)	8	82

In an analysis of all the infiltration tests (n = 375), the Shapiro-Wilk normality test using Statistica (StatSoft, 2010) (normality is rejected for W when p < 0.05) found that infiltration test results (n = 285) when pre-infiltration tests were less than 30 seconds were normally distributed (W=0.99, p=0.064) and pre-infiltration tests that exceeded 30 seconds (n = 90) were not normally distributed (W = 0.87465, p = 0.00000). The pre-infiltration tests that exceeded 30 seconds were categorized with a histogram that implied log-normal distribution (appendices). This implies that once pre-infiltration tests exceed 30 seconds there is typically a precipitous drop off in infiltration rates. A Mann-Whitney U statistical test using Statistica (StatSoft, 2010) expectedly found a difference (p < 0.01) between pre-infiltration testing results.

Discussion

The ASTM method for PICP, ASTM C1781/C1781M – ${18}^{{\epsilon }^1}$ (ASTM, 2029), notes that it is “functionally identical” to ASTM C1701/C1701M as both tests provide “comparable results,” the exception being for “added provisions for positioning and securing the test ring to a discontinuous surface” for C1781/C1781M – ${18}^{{\epsilon }^1}$. As noted previously, testing was conducted by the modified ASTM C1701/C1701M with neoprene instead of non-hardening plumber’s putty with additional strips cut to fill in gaps between pavers to avoid side leakage during infiltration testing.

Kazemi et al. (2017) noted establishing rate of clogging assisted with maintenance to rejuvenate surface infiltration. Drake et al. (2013) noted that even within a single permeable pavement that the effect of maintenance can vary widely. Razzaghmanesh and Beecham (2018) discussed a variety of approaches to maintain permeable pavements, indicating no one approach is “universally preferred” and concluded some of the most practiced methods were “ineffective.” Indeed, Drake and Branford (2013), Winston et al. (2016) and Simpson et al. (2021) discuss the need for multiple passes of equipment and machinery to perform maintenance especially along permeable pavement edges receiving runoff, even noting the need to “in some cases stopping over heavily clogged areas” (Winston et al., 2016).

Much of the literature cited in this paper provide common causes for clogging: watershed area, sources of sediments of contributing area and traffic. However, none of the other studies identified low spots that concentrate flow as a contributing factor. Is the performance of maintenance equipment the variable of concern or is that the accumulation rate of material that clogs permeable pavements is variable as there are low and high spots? This study has shown that the decrease in the infiltration rate along the upgradient edge of the PICP where stormwater runoff enters is highly variable and the lowest spots, as revealed by the HRS, clog faster than other locations along the upgradient. Increasing the number of passes of maintenance equipment or machinery may not uniformly rejuvenate infiltration rates along the edge and more pinpoint maintenance treatments of low spots may be necessary. The high-resolution survey has shown that assumptions of a linear interface between runoff source and permeable treatment surface based on design are not applicable given real-world construction practices and limitations.

Both Table 1 and statistical analysis of pre-infiltration tests would imply the pre-wet test is sufficient alone for identifying clogged areas in PICP. These results validate the previous mentioned conclusions regarding pre-wet tests of Brown and Borst (2014) that pre-wet tests are sufficient for most maintenance requirement determinations. Many more unique infiltrating test locations could be performed in the time it would take to complete the standard two-step infiltration tests.

Conclusions

This project successfully addressed the hypothesis outlined in the Introduction and provides further insight about maintenance for permeable pavements.

The infiltration testing of the original PICP was completed at randomly selected sites and only identified 9 of 204 locations considered to be Clogging (< 5,920 mm/hr) with 8 of these 9 locations 0.1 to 1 m from the upgradient edge. The infiltration testing of the new PICP identified 3 locations as Clogged (<99 mm/hr) and many more locations as Clogging as infiltration testing targeted the edge of pavers with additional locations linearly downgradient of the edge. The larger number of Clogging infiltration tests, 81 of 175, in the second study is attributable to the increased number of infiltration locations along the upgradient edge where clogging was expected to occur, a predicted outcome of the first PICP infiltration study.

Both studies point to gradual surface clogging with a reduction of infiltration rates, as demonstrated by the regression analysis of the mean monthly infiltration rate over time. The new PICP had a narrower gap between pavers than the original PICP; mean monthly infiltration rate regression analysis indicated that the new PICP had a lower initial infiltration rate and steeper decline in infiltration rate than the original PICP which is consistent with previous research (Ehsaei, 2013) where pavers with larger gap sizes had larger infiltration rates and slower clogging. A similar observation was made by Drake and Branford (2013). A stronger coefficient of determination (R²) for the new PICP infiltration testing than the original PICP testing was because of the targeted upgradient infiltration sampling in the latter study, within 2 m (6 ft) of upgradient edge, while the original infiltration testing sites were chosen randomly across the 8.5 m (28 ft) distance of the double-parking rows. The robustness of the results of the newer PICP study which targeted upgradient locations was also demonstrated by multiple regression analysis.

Of all infiltration tests performed in both studies (n = 379), only three sampling locations along the edge and within the diameter of the testing apparatus (0.3 m) were identified as Clogged during the new PICP infiltration study. Further, as shown in Figure 3, results of the new infiltration study at a distance of one diameter of the testing apparatus (0.3 m) indicated twice as many, 11 to 5, infiltration tests were considered Clogging than Infiltrating (> 5,920 mm/hr), while at greater downgradient distances many more infiltration testing locations were considered Infiltrating (72) than Clogging (14).

The results of the original and new PICP studies indicate that even after years of employee and visitor parking on each permeable surface, >2.5 and >3.5 years, respectively, most infiltration test locations away from the upgradient edge predominantly have a condition of Infiltrating or no impact rather than Clogging or impacted. This owes in part to the low ratio of contributing drainage area to pavement width at the EEC. However, Clogging does occur away from the edge, as other factors may be involved (e.g., leaf and plant litter, dirt from vehicles). The predominant factor for Clogged or Clogging locations at the EEC was receiving direct runoff from upgradient sloped driving lanes. Isolated Clogging spots are not going to drive further clogging but progression of clogging through surface runoff from upgradient edge is a measurable effect that can be targeted for maintenance.

Along the edge of pavers, small changes in elevation on the order of millimeters (hundredths of a foot) can concentrate runoff carrying solids to low points resulting in clogging whereas higher points receive less runoff and clog slower. The original PICP infiltration study measured the extent of visible accumulation of solids between pavers. The new PICP infiltration study took the results of a high-resolution survey performed a year and half before the new PICP were placed to predict the location of three locations that eventually clogged. During the study, all sampling locations along the edge eventually are Clogging, even higher elevation spots, while the three locations of concentrating flow were Clogging and then Clogged before the remaining high elevation Infiltrating locations changed to Clogging. Being Clogged, at an infiltration rate of 99 mm/hr (3.9 in/hr) or less, is a more definitive condition than Clogging which can range from 99 mm/hr up to 5,920 mm/hr (233 in/hr).

For PICP applications, a high-resolution survey along with upgradient, targeted infiltration testing may result in more frequent spot maintenance and less frequent area-wide maintenance for the permeable surfaces. The high-resolution survey has shown utility for pinpointing low points of accumulating sediment for the application at the EEC which has a relatively small contributing area and for which cleaning maintenance of the entire permeable surfaces has not yet been necessary. Including a high-resolution survey as part of the as-builts at the conclusion of a permeable pavement construction project would slightly increase construction cost but will provide insight to future maintenance requirements and may reduce future maintenance costs and needs. A high-resolution survey may also be applicable to other stormwater control measures with unplanned, concentrated flow.

In addition to pinpointing infiltration testing locations and performing targeted maintenance of permeable pavement to reduce clogging, porous pavement applications that have widespread and frequent clogging due to larger contributing runoff area should extend maintenance routines to include street cleaning of the paved impermeable contributing areas to reduce source material for clogging.

This study also confirms the previous finding that pre-wet testing for PICP is sufficient to determine Clogging which would help with performing infiltration testing more efficiently for maintenance activities. This will be more important for permeable pavements with larger ratios of contributing drainage area to pavement width than the EEC to determine the downgradient extent of Clogging and maintenance requirements.

Data Availability Statement

Some or all data, models, or code generated or used during the study are available in a repository online (DOI: 10.23719/1520736) in accordance with EPA ORD data retention policies.