FHWA studies new techniques that may be used to estimate head losses through manholes.

Anyone who resides in an urban area in the United States probably lives over an elaborate network of underground pipes that comprise a storm drain system. The system collects storm runoff from streets, parking lots, and other development and conveys it to an outfall.

Approximately every 92 to 183 meters (300 to 600 feet) and at every junction where several pipes intersect, an access hole, commonly referred to as a manhole, is installed for joining pipes with different base elevations, alignments, and/or diameters, and for inspection, maintenance, and repair. Typically the access hole is simply a concrete pipe or box set on its end, with openings for joining pipes into the system, a ladder anchored on one side, and a cast iron cover large enough for a person to enter.

Energy losses in these systems due to friction in the pipes reduce water flows, creating such potential hazards as flooded basements and blown access covers during major floods. Designers analyze these energy losses to select appropriate pipe sizes, set base elevations, and evaluate potential hazards. Estimating energy losses is relatively straightforward, but estimating the losses in the access holes tends to be a challenge because the flow in the junction is so chaotic.

"Most of the energy lost in storm drains is within the pipes themselves, but the energy lost at junctions is not at all insignificant," says Joe Krolak, senior hydraulics engineer with the Federal Highway Administration's (FHWA) Office of Bridge Technology. Ultimately, once the dynamics are better understood, engineers at State departments of transportation, highway planners, and others in the field will be able to apply new methodologies and build improved storm drains with greater cost savings, he adds.

The Original Study

Between 1986 and 1992, FHWA conducted a lab study of energy losses through junction access holes, using relatively large-scale (one-quarter scale) physical models. A preliminary method for determining such losses, based on early results from that study, was published in FHWA's Hydraulics Engineering Circular 22 (HEC 22), Urban Drainage Design Manual (NHI-01-021). A revised method, based on the final results, was coded in the highway drainage HYDRAIN software system.

Practitioners questioned both of these methods when they encountered situations beyond the range of the experimental parameters tested in the lab study. Both methods had limitations when applied to junctions with plunging inflow or with outflow pipes that carry supercritical flow. In addition, both methods are relatively complex and necessitate solutions that may require repeating over time. FHWA plans to update HEC 22 and further develop computer software for storm drain design. The need for consistent technology in FHWA publications and software applications on this subject is urgent. To accommodate that need and overcome some of the difficulties in estimating energy loss in access holes, Krolak initiated an effort to reanalyze the data from the 1986-1992 study.

The Followup Study

Roger Kilgore, principal of Kilgore Consulting and Management in Denver, completed that effort for FHWA in 2005. "The 1986-1992 work improved the methodology for people to improve storm drains," Kilgore says. But Kilgore, who was the principal investigator during part of the 1986-1992 study, subsequently, as a practitioner, has seen the limitations of the existing methods.

(Above) An old, largescale junction loss model is shown here at the FHWA Hydraulics Laboratory. (Below) This smallscale junction loss model, used in the new study, has attached standpipes that measure the hydraulic grade line using vertically mounted contact image sensors.

The new FHWA study separated the 1986-1992 data into two groups. The first group consisted of "base runs" with the simplest configurations (one inflow and one outflow pipe with the same base elevations) to establish first approximations of access hole energy loss. The second group included the more complex configurations used to derive adjustments to the first approximations. The result is a new method, which is somewhat simpler than the existing methods and might improve handling of plunging flow and supercritical flow situations. Concurrently, Krolak requested additional, independent lab experiments to collect data and evaluate the proposed new method.

The original 1986-1992 FHWA lab study used a large-scale model of an access hole with a diameter of 0.6 meters (2.0 feet), which is near prototype for some applications. That original study included 755 test runs. For the current study, researchers decided to use a much smaller scale, easier-to-operate, experimental setup with higher precision instrumentation to investigate a wider range of parameters and study energy inside the access hole itself. In particular, researchers chose to use a particle image velocimetry (PIV) technique to visualize and measure flow patterns. This technique requires special tracer particles to be induced into the flow, which makes the use of a smaller experimental environment much more practical. It also requires transparent models for flow visualization.

The current study has two objectives. The first is to evaluate the proposed procedure, which requires conducting some of the same types of tests that were run in the previous study. But the new tests include a wider range of parameters, such as greater plunge height ratios and steeper pipe slopes. Previous research was limited in that it was applicable to storm drain systems located only in relatively flat areas; the research would not hold up for systems in hilly and mountainous regions of the country, where steep pipe slopes are the norm. Researchers shared data with Kilgore during the study that resulted in several modifications in his proposed procedure.

The second and more challenging objective is to characterize the energy level in an access hole with various inflow and outflow pipe configurations. If that can be accomplished, then the familiar culvert hydraulics analyses can be applied to the access hole that serves as the tail box where inflow pipes enter, and to the head box for outflow pipes where the water exits. Researchers have attempted numerous analyses of PIV data and three-dimensional (3-D) numerical model data, with uneven results. Characterizing energy in the access hole is highly problematic because the flow is so chaotic, and arbitrary assumptions have to be made to obtain results that fall between intuitive limits.

This diagram shows a pair of digital cameras in an angular imaging arrangement and a laser light sheet in the center plane of the access hole. The photo (above) shows an image captured by one of the laser light cameras. Source: FHWA.

Experimental Setup

The new test apparatus for junction energy loss includes three water tanks: a head box, main tank, and tail box. The purpose of the head-box tank is to control the pressure head for the experiments and to allow injection of seeding particles. The junction loss models are mounted inside the main tank, where they are surrounded by still water to minimize distortions for the 3-D and stereoscopic PIV recordings. The main tank also supports a carriage system for two laser distance sensors. These sensors measure the flow depth in standpipes attached to the sides of the inflow and outflow pipes of the junction access hole. This setup is automated to maintain constant flow depth in the access hole during the test run. The water column in the standpipes represents the hydraulic grade line in the inflow and outflow pipes. The tail-box tank is designed to control the tail water.

To accomplish the first objective of evaluating the proposed new method, the researchers performed tests to measure the total loss through the access hole; it was not necessary to measure energy inside the access hole to meet this objective.

The researchers applied two techniques to measure the flow depth. One method used laser displacement sensors pointed at a floating disk in the standpipes. Another method, recently developed at the FHWA Hydraulics Laboratory, uses contact image sensors (CIS) mounted on the sides of the standpipes to scan the water columns. The big advantage of the CIS system is that it can measure all water columns in the standpipes simultaneously, which results in a more precise loss coefficient computation.

Three flow meters provide discharge readings and are used to compute velocity head. CIS sensors pointed at four locations in the access hole measure an average water surface elevation. The models are fabricated from Plexiglas® with a 15-centimeter (6-inch) access hole diameter and 3.8-centimeter (1.5-inch) inflow and outflow pipe diameters, resulting in a relative access hole diameter equal to 4, which is the access hole diameter divided by the outflow pipe diameter.

To meet the second objective of characterizing the energy inside the access hole, the researchers employed 3-D and stereo PIV techniques and 3-D numerical models. The PIV technique is an optical flow diagnostic based on the interaction of light refraction and scattering using nonhomogeneous media. The fluid motion is made visible by tracking the locations of small tracer particles at two instances of time. The particle displacement as a function of time then is used to infer the velocity flow field. The 3-D PIV makes it possible to measure instantaneous velocity flow fields.

A stereo PIV camera system is based on a pair of digital cameras in an angular imaging arrangement. A special geometry is necessary to reconstruct the 3-D field from the two projected, planar displacement fields. Therefore, knowing the precise distance between the two camera lenses and the distance between the cameras and the light sheet is important. The laser system and cameras are mounted on a movable carriage frame to keep the distance constant between cameras and light sheet.

Test Results

The first set of tests, using the new junction loss setup and corresponding small-scale models, was intended to verify scale effects. A subset of the base runs (one inflow and one outflow pipe) was used to analyze scaling issues. The dimensions of the apparatus for the base runs were scaled down by a scaling ratio factor of 1 to 4, and total energy loss across the access hole was measured. Those energy losses, scaled back up to the dimensions of the original apparatus, agreed quite well with the corresponding energy losses from the larger scale experiments. Based on that observation, researchers were confident to proceed with the small-scale models to evaluate the proposed procedures.

Comparing the proposed methodology to the base run tests shows that the new method predicts total losses for the simple one inflow and one outflow pipe very well.

One of the biggest limitations of the existing methods for estimating junction losses was that they did not apply to supercritical flow, where the losses were greatest. Kilgore attempted to solve this problem, but there were no data in the original large-scale experiments to use as a basis for this part of his methodology.

The FHWA lab performed 18 runs in the small-scale experiments, with supercritical flow in the outflow pipe and two inflow pipe configurations at 180 degrees and 90 degrees. The outflow pipe for these runs had a slope of 3 percent. A surprising result of the tests was an almost constant depth in the center of the access hole for a fairly wide range of discharge intensities. Kilgore's methodology does an average job of estimating that depth in the access hole; but it does a good job of estimating the energy grade line elevations for the inflow pipes for supercritical flow situations. Kilgore mentions in his publication that his proposed methodology may not be appropriate for high discharge intensities when flow is into a supercritical outflow pipe, because water might shoot across the access hole in a jet and not expand and contract as expected.

The FHWA researchers conducted 18 runs to model plunge flow conditions. They varied the drop elevation for the inlet between 3 and 10 times the outflow pipe diameter. Kilgore's methodology does a reasonably good job of predicting the influence of plunging flow on the depth in the access hole and estimating the energy grade line elevations for the 180 degree inflow pipe.

This photo shows an averaged velocity distribution vector plot in the center plane perpendicular to the main inflow direction in the access hole.

The FHWA lab also used Kilgore's base runs to explore the kinetic energy distribution in the access hole using PIV. When this aspect of the study is successful, there will be a valid basis for adjusting Kilgore's coefficients even if it does not lead to a full culvert analogy for the energy loss analysis for access holes. The researchers used 3-D numerical model results, as well as PIV physical model results, for the kinetic energy investigation. With the numerical model, depth-averaged kinetic energy could be computed at the very center of the access hole--or integrated over a selected plane or over the entire matrix of vertical node points used in the simulation. For the PIV technique, local velocities in three planes perpendicular to the inflow and outflow centerline were recorded in the access hole. Knowing the total energy loss led to the question of how to separate it into entrance and exit losses. An average kinetic energy was determined in each of the three planes. Regressing these values over the access hole diameter allowed a distribution of the total loss.

The results are very sensitive to the averaging technique used. Depth-averaged kinetic energy generally leads to low numbers because it includes large, ineffective flow areas. To obtain reasonable kinetic energy values from the PIV results, an effective area has to be determined arbitrarily, which is not necessarily transferable to more complex pipe configurations. An attempt using this technique to determine the kinetic energy distribution for a more complicated inflow pipe configuration (pipe at 90 degrees) failed because the flow in the access hole is extremely unorganized, and no zones of effective flow could be detected.

Similarly, the 3-D numerical model gave unexpectedly low kinetic energy values for two of the four base run tests, which would suggest the impossible conclusion that there was a gain in energy as the flow passed from access hole to outflow pipe. One advantage of the numerical model simulation was that no area was arbitrarily disregarded, and if the center point were to be used as the representative energy for the access hole, the simulation might apply to any pipe configuration.

Based on these complications, FHWA is taking a different approach to a methodology for estimating the kinetic energy distribution. The approach will be to relate the contraction loss, as flow goes from the access hole into the outflow pipe, to the contracted area (vena contracta) in the outflow pipe. The 3-D PIV technique can be used to measure the contraction zone in the outflow pipe.

Although it is difficult to measure the kinetic energy in the access hole, it is relatively easy to compute the energy line for the outflow pipe and measure the vena contracta in the pipe using PIV techniques. If the vena contracta is an indirect measure of the contraction loss for the outflow pipe, it would lead to an indirect measure of the total energy in the access hole, regardless of the inflow pipe configuration.

Significant Findings

One concern when starting the new small-scale experiments was the scaling issue. Using base runs confirmed that the small-scale results with precise instrumentation can be used with reasonable confidence to evaluate and develop enhancements to Kilgore's proposed methodology. Small-scale tests allow a much more efficient testing procedure and reduce physical and geometrical constraints.

The proposed new methodology addresses the problem of supercritical flows in outflow pipes. The use of inlet control culvert equations to estimate the initial depth in the access hole for these situations does appear to work very well.

Kilgore proposed a relatively simple equation to compute additional energy loss for plunging flows that accounts for the proportion of the flow that is plunging and the drop height. The proposed procedure is applicable to plunge height ratios, plunge height divided by outlet pipe diameter, up to 10 using the small-scale experimental setup.

Understanding the kinetic energy characterization in the access hole remains the most rational procedure for estimating energy losses in access holes and distributing those losses among several inflow pipes. The two approaches using 3-D PIV and 3-D numerical modeling to analyze the energy level in the access hole did not give satisfactory numbers, due to extremely unorganized flow inside the access hole. Researchers at the FHWA lab now are investigating the more organized flow in the contracted area of the outflow pipe, using the contraction ratio as an indirect measure of the contraction loss in the flow from the access hole to the outflow pipe to back-calculate the energy loss in the access hole.

"It's very worthwhile research," Kilgore says, referring to the recent research at the FHWA lab. "They employed a lot of innovative techniques and newer technology to reduce the scale of the experiments and the time it takes to do them."

That view is bolstered by Krolak, who says that "current methods for estimating the energy loss are unwieldy and overly complicated; some assumptions maybe are too conservative or not conservative enough." The FHWA lab was able to apply new technology to improve the older studies, and "validated some of the old data while making new strides," he adds. The older studies used large-scale physical models. Their size limited what configurations could be considered, Krolak explains. The new technology allows use of much smaller scale models while maintaining (or improving) the accuracy of results, and allowing testing of different configurations and junction types.

Kornel Kerenyi is a hydraulics research engineer in the FHWA Office of Infrastructure Research and Development (R&D). He coordinates hydraulic and hydrological research activities with State and local agencies, academia, and various partners and customers, and co-manages the FHWA Hydraulics Laboratory. He was previously a research engineer at a private company and supervised support staff in the data collection and analysis for this study. Dr. Kerenyi holds a Ph.D. in fluid mechanics and hydraulic steel structures from the Vienna University of Technology in Austria.

J. Sterling Jones is a hydraulics research engineer in the FHWA Office of Infrastructure R&D. He manages the FHWA research program in hydrology and hydraulics and comanages the Hydraulics Laboratory. He provided oversight during the data collection, analysis, and results reporting for this study. He is a registered professional engineer in the Commonwealth of Virginia.

For further information, contact Kornel Kerenyi at 202-493-3142 or kornel.kerenyi@fhwa.dot.gov.