17. Sharp-Crested Weir Construction and Installation

Portable sharp-crested weirs may be used temporarily to provide approximate measurements of small flows in earth channels, lined tunnels, etc. For small earthen channels, the weir may be made from a piece of stiff sheet metal cut in the approximate shape of the cross section of the channel, but somewhat larger with a carefully cut weir notch in the top edge. To set this weir, the metal plate is forced firmly into the soft bottom and sides of the channel normal to the direction of flow. The crest is then adjusted to a level position by tapping down the higher side. Portable long-throated flumes can be used where head is insufficient for sharp-crested operation.

For larger channels and lined tunnels, a weir plate may be installed in a bulkhead that has been sandbagged and sealed in place. The opening in the bulkhead for a weir notch should be cut about 3 in longer than the crest length. This opening will allow insertion of metal crest plates to form the sharp crest and sides of the weir. Some approximate measurements have been made successfully with a combined weir and canvas dam. The bulkhead structure is used to fasten the canvas and bulkhead, which form a dam when held in place across the canal section by piling earth on the lower edge of the canvas.

For simple temporary weirs placed across channels, a flat-topped stake or post may be driven into the bed of the weir pool until its top is at the same elevation as the weir crest. The stake should be located in tranquil water close enough to the channel bank to be accessible. The depth of the water over this post is the head on the crest.

Designing weir structures requires consideration of all the general limits in section 5 and the limits specific to the type of weir crest selected. The approach flow conditions should be as described in chapter 2. In the past, many organizations developed sets of weir box structures for convenience. These weir sets commonly used standard weir equations for fully contracted weirs. However, for economy of structure, these weir boxes skimped on crest height and side contractions and used head-to-crest length ratios of 1 to 1. Some of these preselected weir boxes, when compared with the improved Kindsvater relationships in section 6, indicated errors as high as 20 percent. With today's computers, calculators, and the improved Kindsvater relationships, economy of structure such as smaller structures and less head loss in certain cases can be achieved without loss of accuracy. The only disadvantage is that generating tables is more complex, but the task is easy with computers.

Frequently, suppressed weirs deliver much more water than operators think they have measured. This inaccuracy happens when air vents are not installed, are fully or partially plugged, or are undersized. Suppressed weirs must have proper ventilation of the cavity underneath their nappes. This ventilation is commonly done by installing properly sized pipes in the walls to vent the cavity under the nappe. Another way of providing air is to use the corner of an angle iron pointed upstream in the nappe to spread the water, forming an open airway. Standard equations and tables are valid only when sufficient ventilation is provided. The weir will deliver more water than indicated by the tables and equations when ventilation is inadequate.

This inaccuracy occurs because the nappe sheet seals with the sidewalls, and the falling jet aspirates air from the cavity. The exiting flow carries the aerated water away, causing a negative pressure under the nappe. The negative pressure and some jet backflow raise the water behind the nappe sheet higher than the water exiting just downstream.

The height of pullup behind the nappe depends upon the drop, discharge, and crest length. The height that the water is pulled up behind the nappe is an estimate of the discharge error. For example, if the measuring head on a 3-ft suppressed weir is 1 ft, and the water behind the nappe pulls up 0.3 ft because of air demand, the error of discharge measurement would be about +6.5 percent. If the water was only pulled up 0.1 ft, the error for the same weir and measuring head would be +2.5 percent. However, some of the rise of water behind the nappe is due to backflow from the falling jet.

The design of pipe size to introduce sufficient air depends on the discharge, drop, and the loss of accuracy that is tolerable. Sizing air piping and air vents requires some knowledge of fluid mechanics and is difficult to do. Bos (1989) gives the equations to compute the undernappe pressure and a plot of discharge error versus under-nappe pressure for sizing air vents.

The weir structure should be set in a straight reach of the channel, perpendicular to the line of flow. The weir crest must be level and the bulkhead plumb. Adequate cutoff walls well tamped in place should be used on the weir structure to prevent undermining of the structure. The average width of the approach channel should be set to approximately conform to the size of the box for a distance of 10 to 20 ft upstream for the smaller structures and from 50 to 70 ft or more for the largest structures.

The weir box may accumulate sand and silt to such an extent that discharge measurements will be incorrect. For sluicing silt and sand deposits, an opening may be provided in the weir bulkhead at the floor line beneath the weir notch. This sluiceway should be provided with a suitable cover or gate to prevent leakage. If sediment is a severe problem, then sediment-excluding vortex tubes that bypass bed load with a small continuous flow may be more desirable than inaccuracies resulting from silt and sand.