(Back)
1. Introduction
Whatcom PUD owns and operates two water intakes on the Nooksack River near Ferndale (Figure 1). Plant 1 is downstream of Ferndale and has a maximum intake flow of 50 cubic feet per second (cfs). Plant 2 upstream has a maximum water intake capacity of 28 cfs. Water obtained at each plant is clarified to remove suspended solids (clay, silt, fine sand) prior to pumping to industrial operations near Ferndale. The PUD has a year-round water right for both plants at these maximum flow capacities.
Plant 1 and Plant 2 each have vertical traveling screens (manufactured by FMC Corporation) with 1/8”-mesh. Plant 1 was built in 1965 and Plant 2 was constructed in 1971. Screen installations were considered state-of-the-art when built.
Whatcom PUD believes that the existing Plant 1 screen does not meet current agency criteria for fish protection during low flow periods, when the Nooksack River only submerges about 4’ of the screen. This report includes a preliminary design for replacement of the Plant 1 screen to comply with fish protection criteria published by the Washington Department of Fish and Wildlife (WDFW 1998) and National Marine Fisheries Service (NMFS 1995, NMFS 1996).
Existing vertical traveling screens at the Plant 2 intake were evaluated by Whatcom PUD with respect to current agency criteria for fish protection, and existing screen features were reviewed with WDFW (K. Bates, April 1999). These reviews concluded that the existing Plant 2 screens may meet agency criteria if repaired and retrofitted, but several screen characteristics would need to be further evaluated (side and bottom screen seals, water eddies at varying flow levels, mesh type and opening size).
Whatcom PUD has decided that the Plant 2 screens should be replaced with new screens meeting all agency fish protection criteria, instead of trying to repair and/or retrofit existing traveling screens. Replacement of the existing screens with new cylindrical stainless steel profile bar screens is outlined in this report.
The City of Lynden has a
12 cfs (+/-) water supply intake on the Nooksack River upstream of Ferndale,
consisting of a single large T-screen module and air burst cleaning system.
The proposed design of the new screens for Whatcom PUD’s Plant 1 and
Plant 2 would be somewhat similar to the City of Lynden’s intake, including
stainless steel profile bar screen cylinders, screen orientation parallel with
river flow, screens located near the river bottom, air burst cleaning system,
etc. The City of Lynden’s
T-screen system has performed very well for 10 years, with no significant
problems due to floating debris, screen clogging, silt accumulation in screen,
damage, or corrosion (Klimple 1999).
2. Fish Protection Criteria for New Screens
Criteria for protection of fish at water intakes have been developed by the Washington Department of Fish and Wildlife and National Marine Fisheries Service. Generally, Indian tribes and other agencies also accept these criteria during evaluation and/or design of water intake structures. The most important criteria for screen design include the following (from NMFS 1995, NMFS 1996, and WDFW 1998). It was assumed that the “design fish” for new screens would be salmonid fry (<60 mm length).
General
Structure
Approach Velocity and Sweeping Velocity
Screen Face Material
Civil Works
Operation and Maintenance
3. Nooksack River Hydrology and On-site Rating Curves
The Nooksack River is free-flowing and generally ranges between 1,000 and 10,000 cfs measured at the USGS river gage at Ferndale (Figure 2). The USGS gage is between Plant 1 and Plant 2 on the river (Figure 1); flow records exist from 1967 to present. It was assumed that both plants would have the same hydrologic regime (insignificant inflow or outflow between the plants).
The most important hydrologic and hydraulic information for screen evaluation and design are:
Flow duration (flow exceedance) data to estimate “low design flow” and “high design flow” for screen operation within criteria.
Rating curve at each screen location to show relationship between river level and discharge.
Average water velocity at low design flow, to verify that Vs > Va.
The “low design flow” for screen evaluation was assumed to be 700 cfs based on a review of flow exceedance data for the river (Figure 3). This flow would be the 95% exceedance flow for the lowest flow month (October), approximately equal to a 99% exceedance flow on an annual basis. Proposed new screens would meet agency criteria for this low river level and higher flows.
The “high design flow” for screen operation was assumed to be 6,800 cfs, which is the 10% exceedance flow on an annual basis for the Nooksack River at Ferndale (USGS 1998). This flow will be shown on preliminary design drawings to assist with review of the proposed intake screens. However, the “high design flow” would not control any specific design feature.
Rating curves for the Nooksack River at Plant 1 and Plant 2 are shown in Figure 4 and Figure 5, respectively. Water surface elevation and Nooksack River flow data for the rating curves were collected in late 1998 and early 1999 for this preliminary design.
Average river water
velocities were calculated for Plant 1 and Plant 2 where possible with available
river cross-section data. Calculations
for Plant 1 indicated that water velocity in the river would average 2.4 fps at
700 cfs and 2.7 fps at 1,200 cfs. At
Plant 2, water velocity would average about 1.1 fps at 700 cfs river flow, and
about 4 fps at 6,800 cfs. It is
expected that at higher flows, average water velocity in the river would
increase further at each plant. These
calculations showed that “sweeping velocity” (Vs) would exceed
maximum “approach velocity” (Va = 0.4 fps) at all flows within
the screen design range
4. Preliminary Design for New Screens at Plant 1
This section describes the proposed replacement of the Plant 1 intake screen (vertical traveling screen) with a series of cylinder T-screens aligned along the existing river bank (Figure 6). T-screens would be located in front of the existing intake and in a downstream direction, near the river bottom so they would be submerged at least 6” at the lowest river level (Figure 7). The existing Plant 1 concrete intake structure would be expanded for the new screen design.
Several site features and operational considerations affected design of new screens for Plant 1, in addition to the primary objective of meeting fish passage criteria. These conditions are:
River Bottom Morphology
Since Plant 1 was constructed in 1965, the deepest part of
the Nooksack River has remained near the intake, with the river bottom at
approximately elevation 2’ near the intake (see Figure
7).
However, immediately upstream of the intake, sand deposits to elevation
3’ to 5’ come and go with variations in river flow, riverbed morphology, and
bedload transport. Concrete “flow
training walls” are included in the new screen design (Figure
6) to maintain
“sweeping velocity” along the entire screen structure and minimize the
chance of sand deposits accumulating around new intake screens.
Adequate Submergence of Pump Intakes
Existing Plant 1 pump intakes are located at approximately
elevation 3.5’ (1.5’ above floor level) and require at least 3’
submergence to avoid operational problems (cavitation due to insufficient
suction head). The existing pump
intake bowls cannot be placed closer to the Plant 1 floor, and it is impractical
to lower the concrete floor. Intake
operations would improve during low river flows with a new pump sump lower than
the existing Plant 1 floor level. The
new screen design includes an addition to the Plant 1 intake, with a new bottom
floor at elevation -4’ to reduce problems related to inadequate pump
submergence (Figures 6 and 7).
Intake Structure Size
Hydraulic design of the multiple intake screen modules
required that pump intakes be located along the entire length of the new screen
layout (Figure 6). This requirement
is related to uniform head loss through each screen module (for uniform flow)
and multiple pump intakes. An
intake structure extension about 50’-long is included in the design to spread
the pump intake flow along the proposed screen alignment, and make flow through
each screen as equal as practical.
Major features of the proposed T-screens and Plant 1 addition are explained below. The pump station layout (Figure 6) and cross-sections (Figure 7) are referenced as appropriate.
Each T-screen would be a 28”-diameter cylinder 7’-long with stainless steel profile bars for the screen surface. Openings between bars would be 1.75 mm. Eight T-screen modules are proposed, which would result in a maximum Va slightly less than 0.3 fps. River velocity at Plant 1 averages more than 2 fps at all river flows, providing adequate transport (Vs) for fish moving downstream past the screens.
Screen modules would protrude as little as possible into the river channel, to minimize the potential for damage by floating or submerged debris (Figure 7). Flow entering each screen module would pass through a short section of 18”-diameter pipe and then into the existing and add-on pump station structure. Valves inside the pump station would allow shutoff of each screen for maintenance. Each valve would have a stem and operator extending to the pump station intermediate floor level at approximately elevation 16’ (above high river flows).
Horizontal steel beams attached to the upstream wall of the existing intake structure would deflect large floating debris from the downstream line of screen cylinders (Figure 6). These beams would take the brunt of impact loads on the upstream end of the screen system.
The proposed addition to the Plant 1 intake would be a reinforced concrete structure approximately 50’-long, 18’-wide, and 35’-high. This addition would be attached to the existing Plant 1 concrete walls (Figure 6). The add-on intake structure and flow training walls would convert a riverbank length of about 75’ from steep riprap slope to a vertical concrete wall (Figure 7). A sheetpile cofferdam about 100’-long would extend a short distance out into the Nooksack River for construction of the Plant 1 addition.
Flow training walls would be constructed at the upstream and downstream ends of the expanded pump station to maintain uniform flow velocity and sediment (sand and silt) transport through the new screen installation (Figure 6). These walls would be a cantilevered retaining wall design from elevation -2’ to elevation 14’ (Figure 7). Each training wall would be 12’ long and would have a wingwall extending into the bank. Riprap would be placed around each training wall to transition the new structures into existing riprap banks along this bank of the Nooksack River.
Two holes would need to be cut in the south walls of the existing intake structure to connect the existing pump sump with the add-on intake sump. These holes would be fitted with slide gates to allow isolation of each sump area for maintenance (Figure 6).
One of the existing vertical turbine pumps would be removed because its inlet location would be within 1’ of a proposed screen valve (Figure 6). If left in place, the close proximity of the pump inlet to the valve would result in unacceptably high flow through this one screen module.
Installation of two new vertical turbine pumps is proposed in the add-on area (Figure 6). These pumps would provide more equal flow withdrawal from the common sump chamber, be set deeper in the water to avoid cavitation problems (Figure 7), and would also offset the loss of the pump that needs to be removed from the existing intake. The new pumps would discharge into 16”-diameter pipes leading to a buried 24”-pipe which would connect with the existing water supply pipe system into Plant 1 (Figure 6).
Screen cleaning would be done using an air-burst system supplied by the screen manufacturer. Air burst systems are standard cleaning systems for T-screens and include air nozzles built into each T-screen, connecting piping and valves, receiver tanks holding compressed air, compressor, and control panel. Typically, compressed air is released into each T-screen assembly on a pre-set schedule (timed release), with a manual air release override. The system could also be operated based on pressure difference across the screen, where a certain level of headloss through the screen (caused by debris) would trigger air burst system operation.
Conceptual design of the
air burst cleaning system is based on four independent receiver tanks each
supplying compressed air to two screen modules.
Four sets of piping and valves for air distribution would lead from each
tank to one pair of T-screens. The
system would be controlled to sequence cleaning of each T-screen pair with
cleaning moving in a downstream direction.
The preliminary design size of each receiver (pressure air) tank is
36”-diameter x 6’-high. Limited
floor space in the existing Plant 1 intake structure would require placement of
the receiver tanks in the intake addition (Figure 6).
Final design of the air burst system would be based on the screen
manufacturer’s experience with similar T-screen series installations.
5. Preliminary Design for New Screens at Plant 2
The overall approach for new screens at Plant 2 would be to replace the existing vertical traveling screens with two 44”-diameter x 10’-8” long T-screens mounted on the existing intake’s outside wall (Figures 8 and 9). The “approach velocity” with respect to WDFW fish passage criteria would be about 0.2 fps, well within the 0.4 fps maximum velocity criteria. Components of the screen system are described below.
Each stainless steel T-screen would be mounted on a reinforced steel plate frame (Figure 9). Each frame would be ¾”-thick steel plate reinforced with gussets, with a surrounding flange and compression bolts to seal existing 4’-high by 8’-wide openings in the front concrete wall of Plant 2. A 22.5° pipe bend (30”-dia.) would be welded into the plate frame. T-screens would be bolted to 30”-diameter pipe flanges on the plate frames. Each T-screen assembly would also include a 4”-diameter pipe for compressed air on the downstream side of each pipe elbow (not shown in drawing).
A 30”-diameter gate valve would be bolted to each 22.5° elbow inside the existing intake. Valves would allow the screen to be shut off for plant maintenance, but the valves are not intended for throttling intake flow (flow control is by regulating the intake pumps). A valve stem approximately 33’-high would extend to the intake structure floor level where a gearbox and manual valve operator (wheel) would allow opening and closing each valve (Figure 9).
Four existing openings in the upstream and downstream concrete walls of Plant 2 would be sealed with existing metal slide gates, or new steel plate frames (Figure 9). The existing metal gates need to be examined to verify they will adequately seal the wall openings, and could withstand hydrostatic pressure necessary for intake maintenance. If existing metal gates were not adequate, new steel frames would be fabricated to cover the openings. These frames would be reinforced with gussets and would have a surrounding flange, compression bolts, and neoprene seal (or similar) to make a watertight fit.
The bottom surface of each T-screen would be about 2’ above the existing river bottom. At low river flow (700 cfs), submergence of the top surface of each screen would be about 4’. Each screen would extend about 5.5’ out into the river channel from the existing vertical concrete wall of Plant 2. The upstream screen would be fitted with a conical debris deflector.
Screen cleaning would be done using an air-burst system supplied by the screen manufacturer. Design of the air burst cleaning system is based on two independent receiver tanks each supplying compressed air to one T-screen. Piping and valves (4”-diameter) for air distribution would lead from each tank to each T-screen. The preliminary design size of each receiver (pressure air) tank is 36”-diameter x 8’-high. Each air tank would be mounted on a reinforced steel support frame that would be dropped into a 5’ x 5’ opening vacated by removal of existing traveling screens. Air compressors would be located on the floor near each tank (Figure 9). Final design of the air burst system would be based on the screen manufacturer’s experience with similar T-screen series installations.
Installation of new T-screens at Plant 2 would not require extension of the existing concrete pump station structure or construction of flow training walls, as needed for the new screens for Plant 1 (see previous section). The existing Plant 2 structure layout and pump locations would be suitable for the new stainless steel cylinder screens.
Construction of new screens for Plant
2 would not require a cofferdam in the Nooksack River.
All work would be done using the existing screen structure as a work
platform and by using underwater divers.
6. References
Klimple, T. 1999. Telephone call between T. Klimple (City of Lynden Dept. of Public Works) and P. Tappel (Fisheries Consultants). March 16, 1999.
National Marine Fisheries Service. 1995. Juvenile fish screen criteria. Developed by NMFS Environmental & Technical Services Division, Portland, Oregon. February 16, 1995.
National Marine Fisheries Service. 1996. Juvenile fish screen criteria for pump intakes. Addendum to NMFS juvenile fish screen criteria. Developed by NMFS Environmental & Technical Services Division, Portland, Oregon. May 6, 1996.
U.S. Geological Survey. 1998. Water resources data for Washington. Nooksack River at Ferndale, WA. Gage No. 12213100.
Washington Department of Fish and Wildlife. 1998. Screening requirements for water diversions. WDFW, Olympia, Washington.