
                                  MEMORANDUM

Tetra Tech, Inc.
10306 Eaton Place, Suite 340
Fairfax, VA 22030
phone	703-385-6000
fax	703-385-6007

DATE: 		November 1, 2011

TO:			Paul Shriner and Lisa Biddle, USEPA
	
FROM:		John Sunda and Kelly Meadows, Tetra Tech

SUBJECT: 	Intake/Screen Velocity Compliance Alternatives and Methods for Compliance Determination
Tetra Tech was tasked with contacting vendors to identify techniques (and costs) for determining compliance with intake velocity requirements in the proposed rule for existing facilities.  As described below, there are devices or methods that can be used to measure water velocity, but none are likely to be practical for directly measuring the through-screen velocity.  Direct measurement of velocities in open water or close to a screen is not only possible but is currently conducted on a routine basis at existing water intakes in determining compliance with Federal and State fisheries criteria.  Additionally, there are other technical factors that may affect the compliance options for intake velocity that may be worth considering for the final rule.

I. Direct Measurement of Water Velocity

Water velocity can be directly measured at specific locations in water using several types of devices including propeller-type, electromagnetic, and acoustic Doppler velocimeters.  The propeller-type can only measure velocity in a single dimension (1D) and must be oriented properly.  However, these types of devices are generally not recommended for low velocity conditions in the range of 0.5 fps because any change in resistance to rotation can have a significant impact on accuracy.

Electromagnetic (EM) velocimeters can measure velocity in 1D or two dimensions (2D), depending on the probe type, and rely on measuring the voltage associated with the current that is induced in a conducting fluid as it passes through a magnetic field.  Since metal screens conduct electricity the probe must be kept a minimum distance of 3 inches away from the screen or any metal components (U.S. DOI 2009).

Acoustic Doppler velocimeters (ADVs) measure velocity by analyzing the Doppler shift of sound waves reflected off particulates suspended in the water at a fixed distance in front of the probe and rely on the assumption that the particulates are moving at the same velocity as the water.  The device uses one sound transmitter and three offset receivers, allowing the probe to measure three velocity vectors at once (3D).  Sample volume is relatively small (>1.0cc) and the device's primary limitation is that it is not effective in extremely clear, particle-free water such as spring-fed rivers.

All three of the devices described above can be mounted on a support structure and then positioned at specific locations in front of the screens.

Measuring or Estimating Through-screen Velocity

Tetra Tech contacted a traveling screen vendor regarding methods for measuring through-screen velocity.  The response was that there is really no practical method available for monitoring or direct measurement of through-screen velocities.  While head loss across the screen provides a potential measure, the head loss at through-screen velocities near the proposed 0.5 fps standard would be on the order of several millimeters, which would be difficult to measure with sufficient precision to provide meaningful results.

The through-screen velocity can be calculated using flow data (measured or estimated), total screen panel face area at some specified depth of submergence (water depth), and the percent open area of the screen mesh.  The total screen face area is not the product of the dimensions of the screen but rather takes into consideration the portions of the screen face that are not available for flow.  For example, only about 85% of the area in front of a typical Ristroph screen is actually screen material; the remainder includes the structural components at the sides the screen basket edges, the fish buckets and panel joints, and the foot structure at the bottom.

The simple computational method provides an average velocity over the total open pore space and does not provide any indication of spatial variations across the screen surface.  Such spatial variations could be estimated through use of a computational fluid dynamics (CFD) analysis to model system flow.  However, this would require detailed system design data and would typically cost about $60,000 per screen (one-time cost) to perform (Gathright 2011).  The results would still be estimates, however, and would be influenced by accuracy of the input data and the model itself.

Monitoring Costs

Monitoring Through-screen Velocity

Since there is no readily available method for directly measuring through-screen velocity, velocities would instead be estimated through mathematical calculation as described above.  If the standard involves a simple calculation that does not consider spatial variations (see below for additional discussion), then the costs involved are simply those for engineering services and data gathering.  If these flow estimating methods are deemed sufficient, all of the data necessary should be readily available and necessary calculations can be performed in-house at minimal cost.  If the standard requires consideration of spatial variations, then a CFD analysis may be necessary for each unique screen configuration at a cost of around $60,000 per screen.  In general, there will be no ongoing operations and maintenance (O&M) costs since re-evaluation and additional computations would only occur when significant modifications are made to the system.

Monitoring Approach Velocity or Velocity at Inlets without Screens

The velocity in front of screens or at inlets without screens can be measured directly using several types of velocity measuring devices.  The acoustic Doppler velocimeters appear to be the instrument most commonly used in fish screen evaluations.  An ADV vendor was contacted and indicated that the probe and instrumentation would cost around $9,000 to $11,000 per unit and would last more 10 years (Edelman 2011).  The number of probes and the mounting equipment needed will vary depending on the monitoring requirements and methodology specified.  Additional costs for installing the probe mounting structures will be site-specific but will involve mostly initial costs for labor and installation.  Unless hydrological conditions that vary outside of those during the initial velocity monitoring are expected, follow-up monitoring could be limited in scope.

II. Technical Considerations

As noted above, there are several technical aspects of velocity monitoring that may be worth considering in crafting the requirements for the final rule.  These are discussed below.

Measurement Location (Point of Compliance)

There are two general approaches to creating an intake velocity standard:  setting a limit on the approach velocity or setting a limit on the through-screen velocity. 

One factor to consider in choosing one or the other is the underlying basis of the standard and the behavior of the organisms to be protected.  If free swimming fish and their swim speed is the basis and the response of the fish is to avoid the screen as they are drawn close to it, then the compliance location should be the location where this response occurs.  If the concern is potential injury to fish and non-motile organisms impinged and held against the screen, then the through-screen velocity is a concern since higher velocities are associated with higher pressure differentials across the screen which could promote injury.  For situations where the screen openings are smaller than the fish of concern (including all screens with <=3/8-inch mesh), because the fish are not able to swim within the openings, the velocity of concern is the velocity of the water in front of the screen or the approach velocity.  For screens or devices where the openings are large enough for the fish of concern to swim within the openings, then the velocity of concern is the through-screen velocity.  For intake technologies that rely on a sweeping flow to carry fish and debris across the screen face and away from the screen, limiting the through-screen velocity helps prevent the adherence of debris and non-motile organisms to the screen face.  One advantage of applying the standard to the through-screen velocity is that it ensures that the approach velocity will be equal to (if there is no screen) or lower than the standard.  A disadvantage is that the replacement of a coarse screen that barely meets the through-screen standard with a finer screen could result in the screen no longer meeting the standard even though it is likely that the impact with respect to organisms that were impinged on the coarse mesh screen would be no worse.

An additional concern is the velocity in upstream channels and constrictions.  While a standard based on through-screen velocity will ensure that the approach velocity near the screen will be the same or lower, there is no guarantee that the velocity further upstream in a channel or canal may not be higher.  In fact, many intake screen forebay channels have skimmer or baffle walls, gated openings, or bar racks that may constrict the opening, potentially resulting in velocities greater than the standard.  Sedimentation on the intake channels may also form constrictions.    Thus, in areas where suspended silt and sand are present, even though the channel may have been built or dredged such that average screen velocities are <0.5 fps, solids may be deposited over time upstream of the screen such that the channel velocity may be well above 0.5 fps.

Applying the standard as the through-screen velocity presents an issue regarding whether the application of standard must consider the fact that, as a screen becomes clogged with debris, the through-screen velocity of the remaining open space will increase.

Spatial Variation

In general, the common practice regarding compliance with a through-screen standard has been to apply it as an average across the entire face of the screen or device based on engineering calculations.  For through-screen velocity, this is the only practical method possible; as described above, there is no practical method to measure the velocity within the small spaces of the screen openings.  Whether the point of compliance is within the pores of the screen or at some point upstream, there may be considerable variation in velocity at different locations due to varying hydrodynamic conditions.  The influences of channel walls, pump swirl, obstructions, eddy currents, inlet currents, and distance from different portions of the screen to the pump suction can create significant variations in the velocity at different locations across the screen face.  Fish screening and diversion criteria based on approach velocity often apply the standard to the velocity vector perpendicular to the screen at a distance of 3 inches in front of the screen and may also include a requirement that the velocity be uniform across the face of the screen (U.S. DOI 2009).  If spatial differences are encountered and are a concern, then it may be possible to install baffles downstream of the screen that would help distribute the flow more evenly across the screen face.  Screens with a bypass such as those commonly used in hydro power installations may also have velocity criteria for the bypass velocity vector perpendicular to the screen.

For situations where debris blinding is significant, the degree of debris blinding will affect the through-screen velocity and may also affect the distribution of flow and approach velocity across the face of the screen.  For traveling screens, the cleaned screen panels that are introduced to the through flow (at the bottom for conventional through flow screens) will allow passage of a larger amount flow which will diminish as the screen panels ascend upward and progressively collect debris that impedes passage of water.  Thus, the approach velocity may vary considerably from the bottom to the top of the screen.

The presence of spatial velocity variations raises the issue as to whether the maximum through-screen velocity should consider spatial variations as well.  However, as described above, even if EPA were to establish such a requirement, there would be no direct way to monitor compliance.

If the velocity standard were based on approach velocity, compliance can be determined using direct measurement in most cases and the "maximum" velocity standard could apply to any location in front of the screen.

Potential Concerns with Using Through-Screen Velocity as the Standard

Commenters have raised a number of similar issues, but the list below summarizes several potential issues with the use of through-screen velocity as the basis for existing intakes.  These issues include:

               * There is no available method to directly measure through-screen velocity;
               * Can be estimated using a basic computational method using available data;
               * The maximum velocity at maximum intake flow and minimum water level conditions can be estimated based on historical hydrology data;
               * Unless a CFD analysis is performed, the estimated maximum screen velocity likely will not take into consideration spatial variations of screen velocities at different locations across the screen face and may allow for higher velocities at "hot spots" on the screen.

Potential Concerns with Measuring Velocity at Intakes Without Screens

Issues regarding direct monitoring of approach velocity or the velocity at inlets where there is no screen include:

            * Can be measured directly and accurately using reasonably priced instrumentation;
            * Velocity can be measured in multiple locations and it is possible to measure the true maximum velocity, including those at hot spots;
            * Eliminates the velocity "penalty" associated with the increase in through-screen velocities that occurs when coarse mesh screens are replaced with finer mesh.  The velocity standard should not dissuade use of fine mesh;
            * Is consistent with the rationale that the standard is based on fish swim speed and ability to escape;
            * The evaluation of velocities at minimum low flow can only be performed during low flow events.  For rivers and streams, measurement for compliance at the defined low flow condition may be delayed until the condition occurs;
            * If the proposed numeric standard is applied as the approach velocity instead of through-screen velocity, it would allow some intakes that currently have through-screen velocities that are above 0.5 fps to potentially become compliant since overall approach velocities will be lower than through-screen velocities; 
            * A wealth of existing guidance, technology, and monitoring resources are available since fish screen approach velocity criteria are already being monitored and enforced at many existing fish screens throughout the U.S.

III. Examples of Field Measurements of Intake Velocity

Several publications are available that describe the methods used to perform field evaluations of fish screening facilities that include evaluation of approach velocities.  Attachment 1, "Guidelines for Performing Hydraulic Field Evaluations at Fish Screening Facilities," provides guidance (including photographs) of applications regarding flow measurement technologies and methods.  Typical methods involve mounting the probes on rods or poles in a manner that allows for them to be quickly repositioned for measurement at multiple depths.  Measurements are also taken at multiple locations across the width of the screen.  As noted in Attachment 1, "Mounting instruments in and around structures is always a site-specific challenge.  Devising a workable mounting system for the instrument can take a great deal of time and effort, but it is a worthwhile investment in order to improve the efficiency and accuracy of data collection."  However, once a system is devised, and all equipment and mounting hardware is purchased and installed, the majority of the costs will be associated with the labor required to reposition the probes and collect and analyze the data.  In-house staff can be trained to perform the measurements.

In one evaluation at the Walla Walla River Basin fish screens, water diversion screen measurements were taken using ADV probes at 3 to 5 locations spaced vertically across each screen and at depths of 20% and 80% of the forebay depth (Chamness 2006).  The probes were positioned 3 inches in front of the screens and samples were collected for 30 seconds at a rate of 2 Hz (2 samples per second) for each location.  While many fish approach velocity criteria require measurements to be performed 3 inches in front of the screens, for Ristroph traveling screens with fish buckets, probe positioning must take into consideration the protruding fish bucket and the localized influence on velocity as the bucket passes close to the probe.

In another evaluation at the Roza Fish Screens Facility, approach and sweeping velocity measurements were taken within 3 inches of the screen vertically at 0.05, 0.2, 0.5, 0.8, and 0.9 times the water depth and laterally at 0.5, 2, 4, 6, 8, 10, and 11.5 ft. from the end of a 12 ft. screen using an ADV probe (DeMoyer 2004). 

In another investigation at a hydropower plant of the performance of a prototype 1/8-inch wedge wire fish diversion bar screen with 1/8-inch spaces, the probe was mounted on a modified screen-cleaning sweep bar that allowed for the probe to be selectively positioned at any location both horizontally and vertically across the face of the bar screen (U.S. DOI 2001).  Such an application took advantage of the existence of a cleaning sweep bar and drive track which would not be present at the screens on most facilities.  It was selected because of the screen's position deep within the turbine inlet in a difficult-to-access location.  This situation demonstrates that, if necessary, a sophisticated means of measuring velocity at multiple locations is possible and was justified because they were performing an extensive test on a prototype technology.

At any given time, the screen velocity profile will primarily be dependent on hydraulic conditions such as intake flow, water depth, and screen cleanliness.  Once initial velocity measurements have been collected over a variety of conditions that provide an indication of trends with regard to intake flow, water depth, and screen cleanliness, the frequency of velocity measurements could be reduced.  In cases where significant variations in conditions that are outside of the range that occurred during previous monitoring events can be expected (e.g., intake forebay is prone to sedimentation), more routine velocity monitoring could be required.


References

Chamness, Mickie, Scott Abernethy, Cherylyn Tunnicliffe, "Walla Walla River Basin Fish Screens Evaluations", 2006 Annual Report, Project No. 199601100, 21 electronic pages, (BPA Report DOE/BP-00000652-36)

DeMoyer, Connie. "Roza Fish Screens Facility: Velocity Measurements at a High Canal Flow Rate." Report No. HL-2004-06, Bureau of Reclamation, Water Resources Research Laboratory, Denver, CO. December 2004. 

Edelman, Joel, Sontek/YSI.  Telephone Contact Report regarding cost and application of acoustic Doppler velocimeter equipment for measuring water velocity near water screens. October 27, 2011.

Gathright, Trent. Ovivo. Telephone Contact Report regarding estimating through-screen velocities for traveling screens. October 26, 2011.

U. S. Department of Interior (U. S. DOI). Bureau of Reclamation.  "Guidelines for Performing Hydraulic Field Evaluations at Fish Screening Facilities." April 2009.

