A description of an Acoustic Fluid Level Survey is given below detailing: how the equipment works, what information is recorded, how the information is interpreted, and its practical value & applications.
Example Fluid Level Survey reports:
Example 1: Making lots of gas with a high gaseous fluid level.
Example 2: Pumped off well with pump set below the perforations.
Acoustic Fluid Level Surveys
If you want a quick run-down of what a Fluid Level Gun does, click HERE.
A Fluid Level Gun is a 1-ft long stainless-steel device consisting of a volume chamber, an internal valve, a microphone, and a pressure transducer (analog-to-digital pressure recorder). Here is how the normal acquisition process of a Fluid Level Shot proceeds. The gun is screwed onto the 2” casing valve, and after charging the volume chamber with compressed CO2 (or Nitrogen) to ~200+ psi over the casing pressure, the casing valve in-between the gun and well is opened & the casing valve connecting the well to the flowline is closed: this isolates the well so it is only in communication with the fluid level gun.
Echometer's TWM (Total Well Management) or TAM (Total Asset Management) is the software platform through which the data is acquired and analyzed. Downhole Diagnostic utilizes Echometer's latest wireless equipment and so we acquire the data through the TAM software (and the screenshots given on this page are from the TAM software). After isolating the gun to the well, data acquisition is initiated and the casing pressure and the background noise of the well are recorded for 20-seconds to acquire a baseline. With a click of the mouse the gun is then “fired” as the fast-acting internal valve in the gun is quickly opened—rapidly discharging the compressed CO2 in the pressure chamber—which generates the pressure pulse (acoustic wave) that begins traveling down the tubing/casing annulus.
As the acoustic wave travels downhole and encounters any abrupt changes in the cross-sectional area of the tubing-casing annulus (for example: tubing collars, perforations, TAC's, liner tops, or other obstructions), the cross-sectional changes cause part of the acoustic wave to be reflected back towards surface. These reflections indicate "disturbances" to the acoustic wave and the reflections are picked up and recorded by the sensitive internal microphone in the gun. A plot of the microphone's acoustic recordings is known as the Fluid Level Trace (Fig. 1). Eventually the pressure pulse encounters the top of the liquid level (or some other complete obstruction) and ALL of the remaining acoustic wave is reflected back to the surface microphone (creating the large fluid level “kick” at the right end of the trace). This fluid level kick represents the top of the "gaseous fluid level" (explained later).
Note to reader: the terms LIQUID LEVEL and FLUID LEVEL are used interchangeably here. As my fellow engineers out there will attest, technically speaking the term Liquid Level should only be used when talking about determining the depth to the oil/water. However, the historical industry terminology used for the gun depicted above was a Fluid Level Gun. It is an unfortunate terminology blunder that has taken root and we even find it difficult to not instinctively say Fluid Level at every chance. The key difference to keep in mind is that oil, water, and gas are all fluids—but only oil & water are liquids. Since gas flows up the casing to surface, the equipment described herein is obviously not determining the depth to just any fluid—rather it is measuring the depth to the Liquids. Maybe someday even these old dogs here at Downhole Diagnostic will conform to the proper scientific terminology, but until we do (and the industry as a whole does as well) we are stuck in limbo as both terms are commonly used.
Fig. 1—shows an example fluid level trace as displayed in Echometer's TAM software. A faint backdrop of the well is displayed showing the tubing, casing, perforations, and TAC in relation to the Fluid Level Trace. The red "C" line (Collar) represents the depth to the last tubing collar that the software could pick up (discussed later). The LL (Liquid Level) line is in black and is residing just above the TAC [the TAC is indicated by the triangles]. The down-kick seen at the LL-line is the Fluid Level Kick.
The gun's microphone records the reflected pressure waves in relation to their polarity (UP or DOWN-kicks) and this is of great significance. When the acoustic wave encounters a restriction in cross-sectional area (like a TAC, salt ring, liner top, the fluid level, etc.) a positive pressure reflection reverberates back to the microphone and is plotted on the Acoustic Trace as a "down-kick". Conversely, when the acoustic wave encounters an opening in cross-sectional area (like perforations, a down-tapering from 2-7/8" to 2-3/8" tubing, etc.) a "Rarefaction Wave" (a negative polarity, or reduced-density pressure disturbance) is generated and this disturbance equally transmits through the gas medium back to the surface microphone—and is plotted as an "up-kick". See the bottom of page-4 of our Sucker Rod Pumping Short Course PDF for more examples.
The polarity distinction is significant because it helps to identify what the acoustic wave is encountering along its path and is thus critical for properly interpreting the downhole conditions. For example, if an Up-Kick (opening) is seen at a certain depth: hopefully the well data sheet indicates there is supposed to be an enlargement at that depth (like a set of perforations). If nothing in the tubing or casing records indicate an opening at that point the up-kick could indicate a hole in the tubing or casing and should be further investigated.
The Acoustic Trace is analyzed by the operator to look for anomalies and verify quality data has been acquired. We always recommend operators take a least two Fluid Level Shots while on location. The data signatures of the two shots should correlate and thus provide a means for verifying the data is true and accurate. For example, if a well is pounding fluid the impact load of the fluid pound is often picked up by the microphone as additional background noise—which can sometimes create false up- or down-kicks on the acoustic trace. Turning the pumping unit off and taking a second shot will verify the unusual kicks are real and not erroneous noise. Real downhole markers will not move and should show up at the same place between shots (except the fluid level which can move to some degree due to gas slugging).
After the fluid level reflection has been picked up by the microphone the operator must verify the software package is correctly identifying the tubing collar reflections. Properly delineating the distance between collars is important because the depth to the fluid level is calculated from the Round Trip Travel Time (RTTT) [time from the initial shot till the fluid level reflection returns back to surface] and the average acoustic velocity of the acoustic wave (which is determined by properly identifying the tubing collars).
The acoustic velocity of a gas is a function of the pressure, temperature, and the specific gravity. To determine the precise acoustic velocity for each well, the software determines the acoustic velocity directly from the Acoustic Trace by counting the number of tubing collars seen on a per second basis (jts/sec) and then multiplying that number by the tubing's average length per joint (ft/jt). As seen on the Collar Analysis image below, Echometer's software platform provides different filters through which the acoustic trace can be run to help accentuate the collar kicks so the operator can correctly identify the collars.
Fig. 2.a—illustrates the process where the operator must adjust and verify the tubing collars selected by Echometer's TAM software. The operator must line up the 11-Point Divider (the equally spaced red lines) as best as possible with the collar kicks which the software then uses to compute the average acoustic velocity. Not all of the collars will exactly line up on the red lines due to joints of tubing that deviate longer or shorter than the average length per jt.
Fig. 2.b—shows the same Collar Analysis as displayed above but with a different digital filter being applied to the data to make the collars easier to spot. This is the BP (Band Pass) filter [highlighted in lower left blue box].
Once the tubing collars have been identified, the software then runs the whole acoustic trace through a Band Pass Filter to increase the signal-to-noise ratio of the data. The filter is geared towards highlighting the signal in the frequency interval in which the collars are located. By applying this filter to the data the tubing collars are accentuated so the deeper collars in the well can be identified over the background noise. A plot of all the tubing collars detected in the acoustic trace is shown in Fig. 3 below. The signal from the tubing collar reflections becomes weaker the deeper the acoustic wave travels and eventually the reflections get drowned out by the background noise in the well and can no longer be detected. This point at which the software can no longer pick out the collars is represented by the C-Line shown in Fig. 1.
Anything that improves the signal-to-noise ratio will allow the gun to see further. Possible means of achieving this include: taking a bigger shot (using more pressure or a larger pressure chamber); turning the pumping unit off to reduce noise; or building up the casing pressure by leaving the casing valve closed (the higher pressure increases the ability of the acoustic waves to propagate). If the fluid level is close to surface, however, taking a smaller sized shot often improves the clarity of the data.
As previously mentioned, the acoustic velocity is computed by relating how many tubing collars are counted per second and then relating that number to the average length per joint. As shown in Fig. 3, the software cuts up the acoustic trace into parts and the acoustic velocity is calculated for different intervals along the well (because the acoustic velocity is not constant as it changes as a function of T, P, and specific gravity).
How is the length of tubing below the Collar-Line accounted for, you ask? Well that's a great question, thanks for asking! Bellow the Collar-line Echometer's software merely uses the last computed acoustic velocity right above the C-Line in conjunction with the time duration to determine the length of the tubing below that point. For this reason, it is important to try and make that C-Line as close as possible to the Liquid Level. Ideally, you would like the C-line to be at least 80% of the depth to the fluid level for best accuracy.
Fig. 3—shows the All Collars page in the TAM software. Each strip displays a 2-sec interval with the tubing collars being denoted by the hash marks. The average Jts/Sec for each interval is displayed on the right side for each strip. Note on the bottom strip that only 2 collars are counted before the signal-to-noise ratio degrades to where the collars can no longer be detected (the last marked collar denotes the Collar Line). The squiggly trace continues until the final line at the end of the trace which denotes the depth to the Fluid Level Kick.
Despite all the talk of tubing collar reflections, it is not necessary to use tubing collars to determine the depth to the fluid level (even though it is the most common method). If the tubing is integral joint (doesn't have upset tubing collars) or if the a shot is taken down the tubing so no collars can be seen, there are two additional means of determining the average acoustic velocity so that the depth to the fluid level can be computed:
Acoustic Velocity Method: the average acoustic velocity can be entered straight into the program and used in conjunction with the Round Trip Travel Time (RTTT) to determine the fluid level depth. This can be accomplished by either: a) Computing the acoustic velocity from a compositional analysis of the gas and entering the average downhole temperature and pressure, or b) By using the average acoustic velocity measured from an offset well producing similar gas under similar conditions.
Downhole Marker Method: by assigning a depth to a known downhole marker seen on the acoustic trace (for example, a kick resulting from the TAC) the acoustic velocity can be computed by measuring the RTTT to that marker. The depth to the marker would be taken from the well data records. This method is valuable when tubing collars are not present or the collar reflections are not clear and the C-Line is significantly higher up the hole than the selected Downhole Marker.
Any method that allows the software to compute the acoustic velocity will then allow it to determine the depth to any point on the Acoustic Trace. Direct measurement methods from either the Tubing Collar or Downhole Marker method are preferable, but which method is actually employed just depends on the specific circumstances.
That pretty well sums up the inside scoop on fluid level shot generation, downhole reflections, the recording plotted on the Acoustic Trace, and how Echometer's software uses the acoustic velocity to determine the downhole depths. Now lets move on to the last essential feature of the Liquid Level Survey...
The Casing Pressure Build-Up Test, GFLAP, & BHP's:
Fig. 4—a wellbore schematic illustrating the important values obtained from a Fluid Level Survey (taken from p.1 of our Sucker Rod Pumping Short Course).
Distance to Fluid Level (FL Kick), ft.
Distance to the Dead Fluid Level (depth to FL if all the gas volume were removed), ft.
Gas Free Liquid Above Pump, ft
Pump Intake Pressure, psi
Producing Bottom Hole Pressure, psi
As previously discussed the fluid level "kick" represents the top of the fluid level, AKA the Top of the Gaseous Liquid Level (shown as h_FL in Fig. 4).
However, oil operators are usually more curious to know how much “dead liquid” is residing above the pump and formation. The "dead liquid" height represents the liquid column remaining if all the volume occupied by the gas were removed. The reasons they are more interested in this value is:
Usually, only the liquid column significantly contributes to the hydrostatic backpressure held on the formation.
The downhole pump is only intended to produce liquids (not gasses). The downhole pump can pump gas—as it does when a well is experiencing gas interference—but it does not do it efficiently. And why pump the gas when the gas will naturally flow to surface up the backside? So the dead liquid height above the pump represents how much pumpable liquid the pump can produce.
So even if the Fluid Level Shot indicates the top of the fluid level is 2000' above the pump, the amount of pumpable liquid the pump can produce might only be 100' if the well produces a lot of gas. And knowing this fact makes a huge difference in hydrostatic pressure calculations, interpreting how efficiently the well is drawing down the reservoir, and drawing conclusions on what changes (if any) should be made to optimize the well.
In order to determine the dead liquid height, the volume of gas bubbling up through the liquid column is determined by performing a Casing Pressure Build-Up Test—which is initiated right after the Fluid Level Shot has been taken. Since the casing vale to the flowline is still closed, the gas flowing into the well is bottled-up with nowhere to go—causing the casing pressure to rise. The build-up test records the increase in casing pressure over time (usually a 5-minute interval is sufficient) as shown in Fig. 5.
By knowing the cross-sectional area of the tubing-casing annulus and the distance to the top of the Gaseous Fluid Level, a defined volume is known and this is precisely the volume being pressured up by the influxing gas. Using the Gas Law (PV=znRT) the change in pressure can be used to calculate the volume of influxing gas required to cause the pressure change, and thus the instantaneous gas flow rate (in MCFPD) flowing up the backside can be accurately estimated. This estimate usually correlates with volume of gas metered for that well on recent well tests UNLESS significant quantities of gas are being produced through the pump and tubing (which is not measured by this test).
Fig. 5—the Casing Pressure Build-Up Test as displayed in Echometer's TAM software. In this example, the pressure increased from 54 psi to 65 psi in a little over 7 minutes, which correlates to an average 90 MCFPD producing up the casing. Recent well tests on this well indicated it was making 200 MCFPD, so why the large discrepency? After finding an extremely gassy dyno card and taking a tubing fluid sample that came out as pure foam—the answer became apparant!
Knowing the instantaneous gas flow rate, an estimate of the liquid column gradient (psi/ft) of the gaseous fluid column can be calculated by applying a field derived correlation developed by Echometer. With this value known, hydrostatic calculations can be computed to solve for downhole pressures anywhere along the wellbore by summarizing the surface casing pressure + gas column pressure + gaseous fluid column pressure at the desired depth.
Now applying the information just discussed the two most important quantities derived from a fluid level shot can be quantified:
GFLAP (Gas Free Liquid Above the Pump): this is the "dead liquid" column ("gas-free liquid height") above the pump. A GFLAP of 0' is a completely "pumped off" well, while wells that are experiencing gas interference often maintain a continuous 500-2,000' of GFLAP (yet the pump will only be filling 40-80% full of liquids). The GFLAP represents how much liquid still remains above which the pump can produce.
PBHP (Producing Bottom Hole Pressure): This is the backpressure being held on the formation. The lower the PBHP, the greater the pressure drawdown on the reservoir—and the greater the well's production. This number helps to quantify the Production Efficiency of the well on the reservoir.
Another valuable use of fluid level shots is to acquire the SBHP (Static Bottom Hole Pressure). The SBHP is essentially the composite Reservoir Pressure of the comingled producing formations that are open to the well. For wells producing from tight formations, as most new wells are today that have multiple-stage frac jobs, the SBHP will be somewhat specific to that well's given drainage area. When a well stops producing, the fluid level in the well will rise until the hydrostatic pressure exerted by the fluid column is equal to the average Reservoir Pressure of the open formations—and once it reaches that point, the net fluid inflow into the wellbore ceases (static fluid column).
Since no oil company will unnecessarily turn their wells off for extended periods of time, the best time to acquire a SBHP is after a well has gone down with a failure and right before the pulling unit arrives to repair the well. By taking a static fluid level shot the SBHP can be calculated and used in Vogel's IPR (Inflow Performance Relationship) to evaluate the well's reservoir-drawdown efficiency.
Applying Fluid Level Surveys — Practical Examples:
If the well is pumped-off: it might be time to drop the Seating Nipple depth, reduce the SPM, or verify the run-time is properly callibrated so the pump is not pounding fluid.
If a high fluid-level is found: you might consider increasing the pumping capacity and (if it is unusually high) it could be an indication of a casing leak (run a water analysis) or fluid communication (from an offset frack job, etc.). Additionally, the high fluid level might make the Chemical Company reevaluate the corrosion inhibitor being used or modifying the flush volume.
Doing a dump chemical treatment down the backside? It sure helps in getting that chemical “on spot” when the amount of flush water required to properly displace the chemical to the desired depth is known.
H-15 Tests for the RRC. The Texas Rail Road Commission requires wells more than 25 years old that have been inactive for more than 1-year be tested every year to verify the wells do not pose a potential threat to contaminating fresh water zones. A fluid level shot must indicate the fluid level is lower than the "deepest usable-quality water" zone. See the rules linked here.
Fluid level shots applied for well control. Observing where the fluid level is located before you nipple down the wellhead (or BOP) can be reassuring on those wells that easily get agitated. Or, it can help make sure you don't over-kill the well with kill fluid unneccessarily.
Fluid level shots dramatically help in finding casing leaks, whether that knowledge comes from an abnormally high fluid level or a high anomalous up-kick on the Fluid Level Trace (indicating a potential hole).
Fluid level shots can also be taken down the tubing and, in some circumstances, even with a rod string inside (however paraffin greatly reduces this ability). In certain situations, comparing the fluid balance between the tubing and casing can be valuable. Depending on the circumstances this could help determine if a Packer is leaking (say, on a plunger lift well that has a packer installed due to a casing leak).
On plunger lift wells: A fluid level shot down the casing can be used (in conjunction with the TP/CP) to determine if liquid loading is preventing the plunger from surfacing. Conversely, a fluid level shot down the tubing might find the plunger stuck half way up the tubing string, or the fluid level trace might find an up-kick somewhere above the fluid level—indicating an enlargement at that point (i.e. tubing leak).
The potential applications of Fluid Level Surveys are numerous. The fluid level shot also is invaluable by providing the backdrop from which the Dynamometer Cards are then interpreted.
Continue on the journey my friend! Come, read about Dynamometers and what makes them DynoMight!