The field methodology for drawdown (slug) tests in wells is described in Harvey and others (2000). The formulations of Bouwer (1989) and Bouwer and Rice (1976) were used to determine the hydraulic conductivity for the partially penetrating wells in an unconfined aquifer. The results of the drawdown tests show great variation and are shown in table 11. The lowest hydraulic conductivity was 1.0 ft/d at MP3-A. At this well, the bottom of the screen is at -174 ft NGVD and is in silty sediments. The highest hydraulic conductivity measured at ENR, 242 ft/d, was at MP3-C in fine- to medium-grained sand at an elevation of -47 ft NGVD. The mean hydraulic conductivity of all ENR wells, determined by drawdown tests, was 32 ft/d.

The variation of K_h among wells was greatest in WCA-2A. Site U1 had the lowest and highest hydraulic conductivities in WCA-2A. U1GW4 had a K_h of 32 ft/d at an elevation of -5.8 ft NGVD, and U1GW3 had a K_h of 1261 ft/d at -13.4 ft NGVD. The mean hydraulic conductivity of all the wells in WCA-2A, determined by drawdown tests, was 116 ft/d. The variation in K_h within the same lithologic layers in the study area as well as the minimum, maximum, and mean K_h in each layer are shown in table 11. Three of the layers have a maximum K_h greater by an order of magnitude than the minimum K_h.

The results from the drawdown tests were compared to results from earlier tests in eastern Broward and Palm Beach Counties by Dames and Moore (1988) and Fish (1988), along with Rohrer’s (1999) results from ENR. The hydraulic conductivity values from this study were higher than results from these earlier Surficial aquifer studies. Results from these investigations and this study are shown in table 12.

Hydraulic conductivity values in comparable lithologic units (for example, sand) vary significantly with each study. Dames and Moore (1988) completed slug tests at 11 wells in coastal south Palm Beach County at elevations of -15 ft to -85 ft (1929 NGVD). These wells, completed in poorly sorted sands, have hydraulic conductivity values that range from 1 ft/d to 7 ft/d with a mean of 2.8 ft/d. Slug tests done in Rohrer’s (1999) ENR seepage investigation have hydraulic conductivity values ranging from 0.04 ft/d to 150 ft/d with a mean of approximately 5 ft/d. It should be noted that the wells in the Rohrer study generally were constructed on or near the levees surrounding the ENR. Some of these wells have screens in the unconsolidated levee material, and may not reflect the hydraulic conductivity of the undisturbed aquifer material. This screen placement could affect the average hydraulic conductivity value determined from that study.

This study found the mean horizontal hydraulic conductivity to be approximately 31 ft/d in ENR and 120 ft/d in northern WCA-2A. Sand layers were found to have hydraulic conductivities of 15 ft/d in ENR and 30 ft/d in WCA-2A. Dames and Moore (1988) reported a value of about 3 ft/d. This difference in K further demonstrates the large variation of hydraulic conductivity found in the study areas.

Evidence for Layers Restricting Vertical Flow

Sediment deposition, sedimentary processes, and sediment weathering are important in creating layers in the Surficial aquifer where preferential horizontal flow occurs, or where vertical flow of ground water is restricted. Anthropogenic practices, such as peat removal and blasting for canals, also are locally important. The lithology of the Surficial aquifer in the two study areas is primarily sand with limestone layers concentrated between 10 and -30 ft NGVD. Whereas sands generally are fine grained and well sorted in the ENR, they are coarser and more poorly sorted at S10C in WCA-2A. The presence of the gravel-size fraction changes from not present in the northern part of ENR to gravel in every sample from the southern part of the study area at S10C. The upper sand layer (layer 2) is relatively thin compared to other sand layers or absent, and the limestone layer (layer 3) is thicker in ENR compared with the limestone layer at S10C in WCA-2A.

Core samples include sandy limestone and limestone with varying porosities (11 to 31 percent) and up to 45 percent quartz composition. The sandy limestone primarily was found at approximately -18 ft NGVD. Generally, porosity was controlled by secondary processes. The depositional sequences at the ENR and WCA-2A initially were similar, but subsequent erosion of the upper 60 ft of south Palm Beach County sediments, including WCA-2A, occurred during the transgressive period of the Upper Miocene/Pleistocene epochs. Periods of sea-level regression followed, which deposited new sediments of larger grain size than originally deposited. These new sediments have higher hydraulic conductivity than the older sediments. The limestone formed in and near the WCA-2A is not as hardened by post-depositional processes and is not as vertically restrictive to ground-water flow as the limestone in ENR.

Preferential flow paths in sand generally are in the shallower sand at S10C (layer 2) but distributed at varying depths in the ENR. ENR preferential paths were seen in layers 2, 3, and 4 with the majority in the deeper sands of layer 4 (fig. 4). Results from slug tests show that K values are higher in WCA-2A by approximately an order of magnitude. This difference is greatest in the shallow layer of WCA-2A.

Evidence for restricted vertical flow comes mainly from the ENR site. Dense, low-porosity limestone with low vertical hydraulic conductivity was found at elevations ranging between 3 and -7 ft NGVD in that area. With vertical hydraulic conductivities as low as 0.002 ft/d, those layers locally act as an aquitard. Rohrer (1999) also observed a vertically restrictive layer for flow that he referred to as "caprock." Interpretations made here differ from Rohrer’s interpretations in that this study found the vertical restricting layer is not horizontally continuous enough over the area to serve as the primary layer restricting vertical flow. It also is likely that blasting to create canals and borrow pits produced important avenues of vertical exchange and may have affected hydraulic properties throughout the ENR. The data from interior areas of the ENR show core samples with very high vertical values of K. Also, the return of drilling fluids once was observed to occur 18 ft away from a drill site, at a location far removed from canals or blasting. These observations further support the conclusion that a vertical restricting layer in limestone beneath ENR is not as laterally continuous as the peat layer (with K less than 1 ft/d) that overlies the aquifer. Additional support for this conclusion comes from the Danish Hydraulic Institute (1999), which had to increase vertical K in its simulations of ground-water flow in ENR above what would be expected for a horizontally continuous layer that restricts ground-water flow.

The peat layer, with relatively low vertical hydraulic conductivity on the order of 1 ft/d, most likely functions as the restrictive unit to vertical ground-water flow. Whereas certain limestone layers have vertical K values less than peat by an order of magnitude or more, the horizontal continuity of the peat layer proves its primary importance as the layer that restricts vertical flow.