Content of PetroWiki is intended for personal use only and to supplement, not replace, engineering judgment. SPE disclaims any and all liability for your use of such content. More information
Diverting spinner flowmeter
Diverting-spinner flowmeters are the most accurate of the spinner devices when low total rates and multiphase flows occur. The stream is diverted through the tool’s barrel, thereby raising the velocity of flow and increasing the sensitivity to the point that diverting spinners can detect rates as low as 10 to 15 B/D. A typical 1 11/16-in. tool has a barrel ID of approximately 1.45 in. A flow of 10 B/D results in a velocity of 3.4 ft/min inside the barrel. Because of the limited clearance between the spinner and the barrel, this velocity is enough to overcome friction and turn the spinner.
Furthermore, a flow of 100 B/D passes through the barrel at 34 ft/min, which is sufficient to start the homogenization of the flow, which eventually eliminates phase influence. In casing, a rate of 2,000 B/D is needed to have the same effect around a continuous spinner. Another benefit to multiphase-flow application is that the tool can be calibrated directly for such flow. As long as the diversion is effective, casing size is not a parameter in the calibration, and the tool can be calibrated in the low-rate range where phase bias is still important even inside the tool.
Types of diverting-spinner flowmeters
One type of diverting flowmeter is the hem-packer. The lower end of the diverter contains a coated fabric "skirt" with a 1-in. hem at the bottom. The skirt is opened by the metal struts to which it is attached, so that the momentum of the stream can effect a reasonable seal. A downhole pump inflates the hem with oil contained in the tool body, resulting in a seal that diverts practically the full stream through the barrel and past the spinner. A downhole motor expands and retracts the struts. The response of the tool to single-phase flow is nearly linear. Two interchangeable spinner elements are available—one with a small pitch for high sensitivity at low rates and a second with a larger pitch for rates up to approximately 2,400 RB/D. Because of its fragility, the tool never found widespread use, but is still available at some locations.
Basket (metal-petal) flowmeter
Another type of diverting flowmeter is the basket (metal-petal) flowmeter. On the lower end, the basket is opened by motorized compression of its several struts, each of which is tack-welded on its underside to a corresponding petal cut from sheet metal. As the struts are compressed to open the basket, the petals slip over each other so as to maintain some overlap even when the basket is fully opened. As the basket opens more to accommodate larger casing diameters, overlap between the petals decreases and leakage through the basket increases. Strut travel is limited to prevent excessive opening and leakage. Standard 1 11/16-in. tools accommodate casings as large as 5.5 in. nominal. Some 1 11/16-in. tools can accommodate 7-in. casing. This is an operational feature that should be checked carefully when ordering this service. Also, it is prudent to inquire whether the tool has been calibrated in a section of pipe of diameter close to the intended application. If not, the service company may not know the effect of leakage on spinner response.
At the point where the outside of the struts meets the casing’s inside wall, the metal petals (which are attached to the inside of the struts) cannot deform enough to effect a complete seal. Because of leakage around the struts the low flow-rate response of the tool is nonlinear. Attaching the petals to the outside of the struts would improve the seal, but would wear the petals quickly during use, destroying the tool’s main advantage, its ruggedness.
There is very small clearance between the spinner and the ID of the barrel. This feature assures almost no diversion of flow around the spinner. On the other hand, small particles of debris can plug the tool.
As the spinner turns, it generates a specific number of voltage pulses per revolution. Thus, the pulse rate from the tool can be transmitted through the logging cable for surface recording and determination of the corresponding revolutions per second. The number of pulses per revolution varies considerably from manufacturer to manufacturer. When ordering the service, it is recommended to inquire as to the number of pulses per revolution (more is better).
Typical basket flowmeters are rated for temperatures in the range of 320 to 350°F and for pressures in the range of 15,000 to 20,000 psia. The 1.70-in. tool typically accommodates 3,000 B/D (maximum); the 2.25-in. tool, 5,000 B/D; and the 3-in. tool, 8,000 B/D. In each case, the tool length is approximately 60 in.
Measurements are made with the tool stationary. In a production well, the tool is lowered to the deepest measurement depth and then opened. After recording the measurement at this depth, the tool is pulled up (while open) to the next measurement depth, and so on. When opened in a production well, the tool can be damaged considerably by movement downwards. When considering a diverting flowmeter for an injection well, the user should inquire of the service company whether its tool is capable of such operation.
The chance of having a diverting flowmeter stick in the hole is greater than with a continuous flowmeter. In a sandy flow, for example, the basket recesses may plug, in which case the basket cannot close downhole. If the tool is stuck, the cable can almost always be pulled loose from the cable head and retrieved. If the flowmeter is stuck in the casing, the least expensive approach may be to jar the tool to the bottom of the hole and leave it. If stuck in the tubing, it may be necessary to pull the tubing. Horizontal wells typically have dirt along the bottom of the wellbore; the flow carries the dirt into the diverting flowmeter, which usually plugs.
Diverting flowmeters are easily calibrated in a flow loop, in both single- and multiphase flows. While continuous spinner flow-loop calibrations require many correction factors for their use, such is not the case for the diverting flowmeter.
Only a few companies have extensive flow-loop facilities. If flow loops are not used for calibration, the customary approach is to apply the same calibration to all types of flow or, even worse, to use the same sensitivity coefficient for all situations. Flow-loop facilities at various universities do offer an alternative source of equipment.
Flow loops generally cannot be pressurized or heated, so the calibrations they produce for diverting flowmeters are not exactly correct for multiphase flows, especially gas/liquid. The error, however, is much less than with a corresponding calibration for a continuous flowmeter. Further, some degree of correction is possible for a diverting flowmeter calibration, because diversion of flow around the spinner element is not a problem.
A single-phase calibration of a packer flowmeter in liquid flow is linear, showing constant sensitivity (slope), with no change of the response because of casing size. The corresponding calibration of a basket flowmeter is nonlinear because of leakage around the struts and through the basket. Moreover, the degree of leakage increases with the casing size. For a packer tool in water flow, the RPS-flow rate relation involves a threshold rate (below which the spinner does not turn); for rates above the threshold spin rate, the spin rate increases linearly with rate. For a basket tool in water flow, the initial response (at lower rates) is nonlinear because of leakage around the struts. The final response (at rates approaching maximum) is nonlinear because of leakage through the overlap between petals (i.e., flow through the basket itself). Because of leakage around the struts at lower rates, the basket also has a threshold. Unlike the packer calibration, there is no single sensitivity to the basket flowmeter calibration. The sensitivity is lower at lower rates, then increases with rate until it reaches a maximum at midrange, after which it decreases with increasing rate.
For a packer flowmeter in air at atmospheric pressure, the earlier portion of the tool’s response is nonlinear in gas rate. The lower the gas rate, the less the sensitivity because of the tendency of the gas to "spiral" through the spinner element rather than turn it. With increasing rate, the response becomes linear (constant sensitivity). The value of this sensitivity, however, is approximately one-half the sensitivity in water because of the presence of spiral-velocity components in the low-pressure stream. At normal wellbore pressure, gas density is sufficient to largely eliminate these distortions; however, tool sensitivity remains slightly lower than that for liquid.
In a multiphase flow, the diverting flowmeter should be calibrated at light-phase cuts of 0, 20, 40, 60, 80, and 100%. For each cut, the total flow rate should vary from the minimum the tool can sense up to the maximum the tool can accommodate. This process generates a family of calibration curves, with the total flow rate on the horizontal axis, the light-phase cut as the parameter of the curves, and the RPS response on the vertical axis. The total flow rate should begin at minimum and then include 100, 200, 300, 400, 500, 700, and 1,000 B/D. Above 1,000 B/D, the total flow rate should increase in 500-B/D increments until the maximum is reached. The calibrations should be performed in a pipe size close to the user’s intended application. The deviation angle is unimportant.
When requesting the diverting flowmeter, the user should verify that the service company has a calibration of the tool in the same phases as those that flow downhole in the user’s well, and in a pipe size close to that in the user’s intended application. Unfortunately, very few companies have such calibrations. If the calibration is at a different deviation angle than in the user’s well, this difference is not important.
An example of the use of a diverting flowmeter appears in Fig. 1, which also shows two temperature traces. A 24-hour shut-in temperature log, recorded before the flowing log, is dashed. A flowing temperature log, after 4 hours of flow, is solid. An open set of perforations is shown near the bottom of the figure. Six squeezed perforation sets are above the open set. The stationary measurements from the diverting flowmeter are shown at locations labeled A through D, top-to-bottom.
Above Depth F, the flowing temperature log is initially constant. A water flow of sufficient rate that the flow carries its temperature up with it originates at Depth F. At Depth D, the diverting flowmeter response is 7.1 RPS, showing that the water from F enters the pipe through the open perforations.
At Depth E, the flowing temperature log cools over a short interval, indicating that the water from F mixes with colder water entry at E. The size of the mixing signature at Depth E indicates that the entry accounts for 43% of the total flow, with the remaining 57% produced from the open perforations at Depth F. The flowmeter response at Depth C is 11.7 RPS, also showing that the colder water enters the pipe through squeezed perforations.
Above C, the flowmeter response is 11.5 RPS at Depth B and 11.6 RPS at Depth A. Thus, no more water enters the pipe above C. For a linear response from the diverting flowmeter, 7.1 RPS is 61% of 11.7 RPS; therefore, 61% of the water (approximately 610 BWPD) is from F, with the remaining 39% (approximately 390 BWPD) entering through the squeezed set. This is all the information the diverting flowmeter can yield, because the flowmeter responds only to flow in the pipe. On the basis of this information, one would assume that the 390 BWPD flow comes from the formation at E.
The flowing temperature log, which responds to flow within and behind the pipe, shows that the assumption just made is fallacious. Above depth G, the flowing temperature log is more constant (less slope) than it is below G. Above G the flowing temperature log responds to the full 1,000 B/D water flow inside the casing. At this rate, the stream cools slowly as it moves upward. Immediately below G, the temperature increases more rapidly with depth. This behavior is consistent with an "entry" at depth G that "pulls" the temperature toward its shut-in value. However, the spinner fails to show such an entry at this depth. Instead, the 390 B/D water flow channels down from G and enters the pipe through the squeezed set at Depth E, creating the mixing signature where it mixes with the flow from F. Therefore, the source of the colder water entering through the squeezed set is the formation at G, and not the formation at E. Between Depths E and G, water flows up inside the pipe at 1,000 B/D, while the 390 B/D water flow from G travels down behind the pipe. Because the net upward flow is less than 1,000 B/D and the temperature tool responds to the flow within and behind the pipe, the temperature trace between E and G cools more rapidly with decreasing depth than above G, where the net upward flow is 1,000 B/D.
Noteworthy papers in OnePetro
Use this section to list papers in OnePetro that a reader who wants to learn more should definitely read
Use this section to provide links to relevant material on websites other than PetroWiki and OnePetro