Construction of a New Tollway Below the Groundwater Table
With Nitrate Remediation

A new tollway presently under construction in Orange County, California will extend over 38 km linking three existing freeways. About 2 km of this new roadway construction will occur below the existing groundwater table. The roadway intersects a shallow unconfined aquifer comprised of a highly heterogeneous soil structure, making discharge predictions a challenge. A history of farming in the area has left the shallow aquifer with high TDS and nitrate concentrations, effectively prohibiting the discharge of the groundwater to surface waters and limiting the water's beneficial use opportunities.

An approximately 2500 meter section of State Route 261 is depressed below the existing grade to accommodate noise and aesthetic concerns of the local community as the roadway passed through an urban area. The depth of the roadway varies from the existing grade, to about 7 meters below the existing grade. A shallow unconfined aquifer is located in the area of the roadway alignment with the free water surface about 2 meters below the ground surface (bgs). Preliminary testing indicated the aquifer contained high nitrate and TDS concentrations. Nitrate ranged from 25 to 80 mg/l as N, TDS ranged from about 2100 to 4500 mg/l. The groundwater did not contain any other harmful constituents regulated through the National Pollutant Discharge Elimination System (NPDES) permit process.

The roadway is located in the Newport Bay watershed in Orange County California. Newport Bay is listed by the Regional Water Quality Control Board (RWQCB) as a 303(d) water body. One of the impairments of this water body is excessive fertilization, caused primarily from high nitrates entering the system via San Diego Creek, the primary stream tributary to the Bay. San Diego Creek is ephemeral, though the typical base flow is from 0.3 to 0.4 m³/s during dry weather. Dry weather ambient nitrate concentration in the Creek is about 16 mg/l as N.

Discharge of the groundwater to the adjacent Creek and ultimately Newport Bay was prohibited at the start of the project by the RWQCB due to the existing fertilization problem in the Bay. The existing basin plan has a nitrate objective of 13 mg/l which was subsequently recognized by the Board as "not protective". Groundwater would have to be treated by the Contractor to be discharged to surface waters.

State Route (SR) 261 is being constructed as a design build project. The Contractor was awarded the project in June of 1995. The tollway, consisting of portions of SR 261, SR 241 and SR 133, must be completed per the contract by December 1999. Consequently, delays in construction could result in costly liquidated damages to the Contractor.

The shallow unconfined aquifer was investigated to understand with more accuracy the hydraulic grade line and the flow rates that could be expected during construction and over the long term. Eleven continuous cores were completed as 1- and 2-inch piezometers. The results served as verification that shallow groundwater occurred consistently along this portion of the tollway. Further investigation included pumping tests to evaluate hydraulic parameters and an additional deep boring to obtain more comprehensive lithologic information.

The pumping tests indicated that the shallow aquifer was essentially a single unconfined anisotropic unit. The average horizontal hydraulic conductivity was determined to be about 6.1 m/day. The results of the field investigations were used to construct a finite element groundwater flow model to determine long-term flow rates assuming a passive subdrain system beneath the roadway, and to estimate the lateral extent of drawdown.

The roadway would be constructed an average of about 7 meters below the existing grade, groundwater was generally located about 3 meters below the ground surface (bgs). Construction dewatering would be achieved through a combination of passive and active dewatering systems. Initial discharge estimates for an active-only system were about 11,000 liters per minute, and as low as 5,600 liters per minute for a passive dewatering system.

Several alternatives for disposal of the groundwater during construction were investigated. On-site treatment was considered, but skid-mounted units would be costly to purchase or lease, take a significant lead-time to ship to the site, and would not be highly portable. It was determined that disposal to the sewer system during construction was the only alternative that would not impact the project schedule.

An agreement was reached with the local water district as well as the Sanitation District that operated the treatment works. Connections to the sewer system involve a capital facilities charge and a service fee. Charges for temporary connections can run as high as $0.4 per cubic meter. The Contractor entered into an agreement with the local water district, which in turn entered into an agreement with the Sanitation District for service for the project. Since the flow did not contain BOD or suspended solids, a 'flow only' charge base was developed and capital facilities were leased on a per year basis. The Contractor currently pays about $0.19 per cubic meter to dispose of the groundwater to the sewer system.

The regional water wholesale District in the area also levies a Basin Equity Assessment (BEA) and Replenishment Assessment (RA) for groundwater extracted within the District boundaries. The District Act allows for exceptions to the assessments provided the applicant can demonstrate that the water is not suitable for domestic use or irrigation, and that it does not recharge an aquifer used for either of these purposes. The Contractor successfully demonstrated that: 1) the high nitrate and TDS content of the groundwater precluded both domestic and irrigation use of the water; and 2) the shallow unconfined aquifer was separated from the lower principal aquifer by an aquitard. As a result, the Contractor avoided charges of $0.07 per cubic meter and an additional $0.28 per cubic meter on 25% of the total volume extracted.

The post-construction disposal of the groundwater posed a significant problem. The sewer system was understood, by agreement to be a temporary solution. Long-term disposal of the groundwater to the sewer system was infeasible for several reasons. First, the cost was very high, much higher than onsite treatment. Second, the groundwater contained high TDS. The local water District reclaims much of the effluent in the district. The District would not allow the long-term disposal of groundwater with such high TDS in their system. Finally, both the RWQCB and the local sewer Districts expressed their desire for the Contractor to find a long-term beneficial use for the groundwater.

A significant number of alternatives were investigated for use and/or disposal of the effluent over the long term. Table 1 lists the alternatives, which will be described in reverse order:

Alternative 9 involved terminating the roadway prior to the aquifer location and about 2.4 km short of the planned location. This alternative was not feasible for two main reasons. First, the implications relative to traffic circulation were serious, by terminating the roadway prior to its planned location, circulation patterns would be compromised. Second, the local communities would not allow such a drastic change to the roadway configuration.

Alternative 8 would construct an impervious watertight roadway section eliminating the need for lowering the water table. The capital cost of this alternative was extremely high; it was estimated that about 3000 piles would be required to keep the section from floating, and the invert slab would be about 0.6 m thick. Further, the long-term maintenance and performance of the structure over such a great distance (2500 meters) was suspect.

Alternative 7 was a proposal to pump the groundwater out of the watershed to a neighboring watershed where nitrate was not a problem. The construction of a pipeline about 9100 m long would be required through a highly urbanized area. The right-of-way acquisition alone would exceed cost and schedule limitations.

Alternative 6 involved the construction of a reinjection well field. In general, reinjection of flow is likely to be difficult due to the low reinjection rate that is expected in the local area (from 19 liters per minute to 38 liters per minute per well). This would require a well field of up to 70 wells and an area of over 8 hectares (ha). Right-of-way costs would be high, as would the ongoing maintenance of a large well field.

Alternative 5 would discharge the groundwater to the brine line of a regional desalter facility. The regional reverse osmosis (RO) plant is planned as a groundwater cleanup project for the nearby El Toro Marine Corps Air Station. This alternative would be a good solution as the brine line for the plant was planned to pass near the Tollway. However, the construction schedule of the RO plant is sufficiently uncertain as to eliminate this alternative from further consideration.

Alternative 4 was a proposal to discharge the groundwater to the local creek, and intercept the flow further downstream through a diversion structure. The discharge would be diverted to a wetland area where denitrification would occur in a managed system. Preliminary studies of this system indicated that removal in the pond system would not be adequate due to limited land availability.

Alternative 3 would pump the groundwater to an existing RO plant used to produce domestic water. Concerns relative to the safety of the water supply were the primary reason this alternative was not pursued. The groundwater would be withdrawn through the roadway subdrain system with numerous locations for maintenance and cleanout. Such a system was not consistent with the normal wellhead development protocols. In addition, about 15,000 feet of pipeline would need to be constructed along with associated pumping facilities.

Alternative 2 would construct advanced treatment on site, in the form of a reverse osmosis plant to create either a potable water source or a water source suitable for irrigation. Income from selling the resulting water supply could offset the plant operating costs. It was concluded that the groundwater could be treated using an RO process with about an 80 percent recovery. Cost to produce the product water was estimated at about $0.4 per cubic meter, exclusive of BEA and RA assessments that would also be levied. Such costs would push the finish water cost to about $0.53 per cubic meter. Wholesale cost of irrigation water in the area is about $0.20 per cubic meter, making the alternative unattractive. Domestic water in the area wholesales for about $0.36 per cubic meter.

Alternative 1 was determined to be the best solution. Initially, ion exchange was planned but the process creates salt brine that must be disposed of to the sewer. The local sewer agency would not agree to this process due to the problem with increasing TDS in the system. Consequently, a biological process was selected. The biological process would incorporate denitrification using bacteria and a methanol feed as a carbon source. Filter backwash would consist of ambient TDS and process introduced BOD only, nitrates in the groundwater would be converted to N₂ gas.

An NPDES permit was secured from the RWQCB to operate the plant. Final effluent limitations included a 13 mg/l total nitrogen and a 20/20 standard for BOD and TSS. This allowed the filter backwash to be discharged to the creek rather than the sewer system saving significant O&M cost. Not only would normal capital costs for sewer service be required, but reclaimed water would be necessary for filter backwash to keep TDS discharge to the sewer system within acceptable levels.

The estimated operation and maintenance cost for the plant is about $100,000 per year, including capital recovery. This is equivalent to about $0.14 per cubic meter of water treated.

There was exceptional interest in finding a beneficial use for the groundwater encountered during the construction of this tollway. About 550 acre-feet per year will be produced by the dewatering system. Domestic and agricultural uses were precluded due to the high TDS (2500 to 3500 mg/l) in the groundwater. Reverse osmosis was determined to be one of the only technologies available for removing TDS to produce a product water suitable for retail use. The cost of producing water with a TDS of about 700 mg/l was determined to be about $0.38 per cubic meter, exceeding the local cost of both reclaimed and domestic water at the wholesale level.

Water is a finite resource in California but its efficient management and use remains an elusive goal in many instances as this case suggests. Contaminated groundwater must be viewed as a resource rather than a waste product if sustainable development is to be achieved.