Until recently, all power and potable water for the Island of Curaao, Netherlands Antilles, was produced by Aqualectra Production at its facility at Mundo Nobo, about 2 km (1.25 miles) west of Willemstad, Curaao's capital. This changed as a result of Aqualectra's decision to:

• Reduce its dependence on thermal desalination processes.  Aqualectra had been concerned about the rising cost of fuel for its thermal processes and the effects on its water tariff.
• Decentralize its thermal and reverse osmosis water treatment capacity from the existing location at Mundo Nobo.  It had been determined that all of Aqualectra's water and power production capacity at Mundo Nobo would be relocated elsewhere on the island to make way for new tourist infrastructure.

Aqualectra's first step was to commission Suez Degrmont to construct the 18,000 m³/day or 2,500 m³/hr (11,000 gallons per minute [gpm]) Santa Barbara Seawater Reverse Osmosis (SWRO) Plant adjacent to Fuikbaai, a Caribbean inlet protected by a coral reef.  This site is approximately 9 km (5.6 miles) to the east of Willemstad (Figure 1).  PB served as owner's engineer responsible for providing oversight and input as necessary through the planning, design and construction phases and conducting quality control checks.

Figure 1: Location map.

Figure 2: A rectangular pit for the seawater pump house was excavated with sheet piles driven into the coral limestone.

Figure 3: George Schlutermann of PB and a Suez Degrémont employee estimate flow from the discharge pipes during the long-termpumping test of the pit.

Figure 4: Water level elevation versus time.

Figure 5: Salinity.

Original Plan Presents Two Primary Challenges

The original plan was to construct a seawater pump house and use a seawater intake as the source of supply.  This plan required installing a high density polyethylene (HDPE) conduit pipe across Fuikbaai, through an inlet channel, across an environmentally sensitive reef, and down the sea slope.  The intake was to be located about 600 m (2,000 feet) off shore.

Dewatering Challenge.  An 8-m (26-foot)-deep pit was excavated in the subsoil to construct the seawater pump house.  For dewatering purposes, sheet piles were installed in an 8-m by 17-m (26-foot by 56-foot) rectangular arrangement (Figure 2).  The large inflow of groundwater into the sheet pile area led the contractor to believe, however, that dewatering the pit would not be possible with conventional means.  (A 700 m³/hour (3,000 gpm) pump was tried.)

Permitting Challenge.  Difficulty arose in obtaining permits needed for the source of supply and for residuals disposal.  Normally, permits are in place before a project is released for bidding.  The client and the contractor had assumed that permitting for the designed intake and outfall lines would be routine, however, so construction began concurrent with the permitting efforts.  But then, two adjacent property owners and six government agencies objected to the permitting associated with the intake line.

Innovative Alternative is Proposed and Tested

Rather than construct a pumping station within the pit and the planned intake line, Suez Degrmont and Aqualectra proposed using the pit as the seawater intake source.  If feasible, this approach would resolve the dewatering problems and the permitting problems.

A large-scale pumping test was required to determine if the required amount of water could be obtained using this alternative.  The test was planned to start on September 21, 2004, but was abandoned a week later because the pumping capacity of 3,500 m3/hour (15,500 gpm)-the test design maximum flow rate-could not be reached with the pumps we had available.  Only a few large-capacity pumps are available on Curaao.

While we waited for pumps that met this capacity to be installed and working properly, we prepared a testing program. To get ready for the test we:

• Cut several holes in the sheet piles to allow free draining of water into and out of the sheet pile pit area
• Established five water level observation points
• Marked these points with red paint, surveyed their heights, and measured distances relative to the seafront.

The pumping test started on October 18, 2004. Water level measurements and water quality samples were taken 24 hours-a-day. Pumping was performed in four steps:

• 1200 m³/hour (5,500 gpm)
• 2500 m³/hour (11,000 gpm) 3100 m3/hour (13,500 gpm)
• 3500 m³/hour (15,500 gpm).

After the pump test concluded on November 2, 2004, we carried out a full-scale recovery test until November 7, 2004.  Afterward, a long-term pumping test was conducted at the initial design rate of 2,500 m³/hour (11,000 gpm), the current capacity of the facility, to monitor water quality on a long-term basis (Figure 3).

After the conclusion of the pump testing, we developed a three-dimensional numerical groundwater flow model using the finite-difference MODFLOW code of the U.S. Geological Survey to represent conditions in the vicinity of the test pit.  (A numerical model was used due to the non-homogeneous lithology of the site vicinity.) 

Each pumping rate was operated for a period long enough for steady-state conditions to have been reached (Figure 4).  We used a calibrated flow model to perform various simulations for testing the effects of the different flow conditions on the drawdown inside the pit.  Two of the simulations included lining the pit floor with concrete, which was a possible final design option of the proposed seawater intake pit at the time.

The results of the theoretical modeling led us to conclude that:

• The pit would provide sufficient water for all operational pumping rates equal to or lower than the ultimate design flow rate of 2,500 m³/ hour (11,000 gpm), and the maximum test pumping rate of 3,500 m³/hour (15,400 gpm), even considering tidal fluctuations.
• The test pit should be sufficient to provide water for the projected 20-year life with the exception of any unforeseen or out of historically observed conditions.  In a worst-case scenario, where water levels could be as much as 0.6 m (2 feet) lower than those simulated in the modeling due to tidal effects, the total drawdown in the pit would  be above the pit bottom but might be too close to the pump intake elevation to be operationally feasible.
• The inflow coming through the bottom of the pit was considerably less than the horizontal flow coming through the walls.  (This conclusion was based on the fact that when we assumed a concrete lined pit bottom, there was no unacceptable drawdown in the test pit.)

Tests Determine if Fresh Water Is Drawn In

We believed that the water being withdrawn from the pit was predominately seawater with a direct connection to the sea, basing our statement on background water levels, tidal fluctuations, pressure gradient variations; and the aerial extent of the cone of depression during pumping.  Nevertheless, Suez Degrmont analyzed water quality at the test pit and its salinity was compared to that of sea water to determine if fresh groundwater was being drawn in.  Salinity was calculated as the field-observed conductivity measurement times a conversion factor of 640 mg-cm/mS-mg.  It is important to note that conductivity measurements were taken under static conditions (no pumping) and during pumped conditions throughout the pump test.  Hypothesis testing was performed to determine if the water quality in the test pit was statistically different than the water quality in the ocean:

• Results of an F-test showed that the standard deviations or variances of the pit water quality data and the ocean water quality data were not statistically different.
• Results of a t-test showed that the mean of the pit water quality data and the ocean water quality data were not statistically different. 

The salinity of the water collected at the test pit and the ocean during the pumping test and during the static period after the pumping test is illustrated in Figure 5.  Based on observation of these data and the statistical hypothesis testing, freshwater is not being drawn into the pit.

Added Benefits

Water quality analyses indicated that the water quality delivered to the plant's pretreatment system far exceeded that expected from the originally planned seawater intake.  Seawater in the area had a silt density index (SDI) range of 5 to 10, whereas the permeable pit had an average SDI of 1.2, indicating an excellent raw water quality.  As a result, very little chemical pretreatment is required.  An SDI ranging from 0.2 to 0.7 is achieved from the cartridge filter vessels and is applied to the membranes.

In summary, the potential advantages of a pit intake source versus a conventional ocean pipe intake include:

Figure 6: Rebar reinforced concrete panels constructed with holes replaced the original impermeable sheet piles.

Figure 7: Completed pit.

• Reduced environmental impacts
• Superior water quality to the pretreatment system in terms of:
  - Less particulate matter
  - Better turbidity values
  - Lower SDI
  - Less biological growth
  - Reduced risk of ship-oil fouling or cruise ship sewage contamination
  - Lower chemical, power and labor usage
• Reduced conflict with offshore recreation
• Less risk of pipeline damage due to dredging or ship anchorage
• Overall lower operations, maintenance and construction costs.

The construction of the final pit and seawater pump station was completed in July, 2005 with follow-up performance testing in the fall of 2005.  As a result of the original sheet piles having had holes drilled in them, in the final construction the sheet piles were replaced with rebar reinforced concrete with holes, as illustrated in Figure 6.  The completed pit is shown in Figure 7.  The use of the pit commenced in early December, 2005 with official inauguration of the plant occurring on January 19, 2006.