4. Submergence

All flumes have a minimum needed head loss to assure that free flow exists and that only an upstream head measurement is needed to determine discharge rate. This required head loss is usually expressed as a submergence limit defined by the ratio of the downstream head to the upstream head, both referenced to the flume throat bottom. The term "modular limit" is defined as this limiting submergence ratio for a particular flow module, which causes no more than a 1-percent deviation in the upstream head reading for a given discharge. When these limits are exceeded, an additional downstream head measurement is sometimes used to extend the measurement range of a flume, particularly for Parshall and cutthroat flumes, but at considerable loss of accuracy. Submergence also increases upstream channel depth, decreasing the upstream velocity, which may aggravate sedimentation problems.

Long-throated flumes can tolerate high submergence in some cases. Trying to extend their measurement range with a downstream head measurement is not recommended. They can be designed to have submergence limits (modular limits) ranging from 65 to 95 percent, depending on discharge rate, shape, and exit channel energy conditions. For example, a flume discharging into a channel that is similar in size and shape to the approach channel can have submergence limits that calculate to exceed 82 to 95 percent for minimum to maximum flow rate, provided an expansion section is used, and from about 72 to 93 percent without an expansion section. The same flume, when discharging into a lake, may have submergence limits of only 65 to 80 percent, decreasing further to about 60 to 70 percent if there is no expansion section. Thus, some knowledge of the installation site is needed before a required head loss can be assigned.

Visual determination of limiting submergence for most flumes can be difficult. However, for long-throated flumes, this condition is relatively easy to recognize. Several references offer guidelines in terms of standing diagonal wave locations to aid visual deter-mination of flow submergence (Bos et al., 1991; Clemmens et al., 1993). In general, if the downstream hydraulic jump causes a frothy wave line across the channel that is at or beyond the end of the contracted throat section, the flume has not reached its limiting submergenceCthe modular limit. If the wave is on the throat, or no wave is visible, the flume is beyond its submergence limit, and the measurement would be invalid. For long-throated flumes, painting lines projecting up the channel walls to denote the downstream end of the crest to aid visual distinction is suggested (Bos et al., 1991). Distinction in terms of wave location in Parshall flumes is not so clear. In the absence of visual observation in any of the flumes, automatic recording operations may need a second downstream head measurement if the opportunity for excessive backwater exists. This measurement would warn of invalid data.

Some states have laws that require Parshall flumes by name for certain situations. Past designs for Parshall flumes tended to overuse submergence for economic savings with the only caveat that submergence should not exceed 95 percent. However, today, designing for this level of submergence is not considered good practice in view of accuracy loss and a hysteresis discontinuity in the submergence correction function described in the section on Parshall flumes in this chapter.