Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Astronomical spectroscopy wikipedia , lookup
Gaseous detection device wikipedia , lookup
Ellipsometry wikipedia , lookup
Anti-reflective coating wikipedia , lookup
Optical coherence tomography wikipedia , lookup
Retroreflector wikipedia , lookup
Photon scanning microscopy wikipedia , lookup
Atmospheric optics wikipedia , lookup
X-ray fluorescence wikipedia , lookup
Downhole Optical Analysis of Formation Fluids Rob Badry Calgary, Alberta, Canada Derrel Fincher Sugar Land, Texas Oliver Mullins Bob Schroeder Ridgefield, Connecticut, USA (FT) Formation (RFT) Repeat Tester Formation Tester 1955 1975 Tony Smits Fuchinobe, Japan (MDT) Modular Formation Dynamics Tester 1992 In the past, wireline formation samplers have not been able to see the fluid they were sampling. Electric power module Downhole optical analysis of fluid before sampling removes the Hydraulic power module blindfold to reveal oil, water or gas. The sample chamber needs to be opened only when the Probe module desired fluid is present. Bringing formation fluid samples to the surface for examination was a novel wireline advance when it was introduced in the early 1950s (right ). Run in open hole or cased hole, the Formation Tester (FT) took a sample of formation fluid where analysis of earlier runs of resistivity and porosity logs showed promising zones. The FT consisted of a sealing packer and probe system that could be set against the formation. Once this was set and opened, formation fluid drained into a sample chamber. The entire sampling operation, from set to retract, was monitored using a pressure gauge. The sample chamber was closed only when pressure stopped increasing—implying the chamber was full and at formation pressure.1 For help in preparation of this article, thanks to Hifzi Ardic, Schlumberger Wireline & Testing, Montrouge, France and Robert Gabb, Schlumberger Wireline & Testing, Livingston, Scotland. MDT (Modular Formation Dynamics Tester), OFA (Optical Fluid Analyzer) and RFT (Repeat Formation Tester) are marks of Schlumberger. 1. Finklea EE: “Use of Pressure Buildup Curves from the Formation Tester for the Evaluation of Permeability and Reservoir Pressure,” The Technical Review 9, no. 3 (August 1958): 30-35. January 1994 Dual-probe module Flow control module OFA Optical Fluid Analyzer module Multisample module Dualpacker module nEvolution of wireline formation testers. The Formation Tester (FT) had a pressure gauge to monitor sampling into a single chamber. Pretest chambers introduced with the RFT Repeat Formation Tester tool allowed a check on seal integrity and gave an indication of permeability before sampling into one of two chambers. The introduction of the dualpacker module, with the MDT Modular Formation Dynamics Tester tool, allows sampling when seal failure might be a problem. The pumpout module is used along with the OFA Optical Fluid Analyzer module to confirm the presence of desired formation fluid before one of several sample chambers is opened. Sample module Sample module Pumpout module 21 The FT’s probe and packer could be set only once per trip in the hole. This created a couple of problems. If the formation has low permeability, the sample chamber could take hours to fill, delaying rig operations and increasing the risk of the tool becoming stuck. Sampling in low-permeability formations was therefore often aborted. But sampling also had to be aborted if the seal between packer and borehole wall failed, indicated by a sudden increase in sampling pressure to hydrostatic. The only remedy was to pull out of the hole, redress the tool and try again. The next generation of testers addressed these difficulties. The RFT Repeat Formation Tester tool, introduced in the second half of the 1970s, allowed an unlimited number of settings or pretests before sampling was attempted. Pretest chambers were used to indicate the permeability and to check for seal failures. During a pretest two small volume chambers opened producing pressure drawdowns. Knowing the amount of drawdown for each chamber gave two estimates of permeability. Once the pretest chambers were filled, formation permeability could also be calculated from the subsequent buildup to formation pressure. 2 Sudden increase to hydrostatic pressure during a pretest showed seal failure. Testing the formation first allowed sampling to be carried out in zones where seal failures did not occur and where permeabilities were high enough to allow one of two sample chambers to be filled in a reasonable amount of time. However, RFT samples suffered important limitations: the sample too often contained a large percentage of mud filtrate and the flowing pressure sometimes dropped below bubblepoint changing the sample characteristics. Even when the sample was formation fluid, it could have been water or gas and of no interest to the oil company. These drawbacks frequently led to expensive and timeconsuming operations, to say nothing of frustration caused by the absence of anticipated information or the presence of data later found to be useless. The latest generation formation tester, the MDT Modular Formation Dynamics Tester tool, overcomes these problems.3 22 The MDT Tool How the OFA module works In the MDT tool, unwanted fluid is expelled from the tool using the pumpout module. During sampling, the engineer can monitor the resistivity and temperature of fluid in the flowline while pumping it directly into the borehole or into a dump chamber. When fluid quality is judged to be representative of the reservoir, the pump is stopped and pure formation fluid can be diverted to the sample chamber or, if a sample is not required—often the case when formation water or gas is indicated—another zone can be tested. To prevent gas from coming out of solution during sampling, pressure is maintained above bubblepoint using throttle valves in the sample chambers controlled by surface software. Maintaining pressure above bubblepoint reduces drawdown, which helps prevent crumbling in soft formations. Excessive drawdown can result in seal failure and hence mud contamination of a sample. Drawdown can also be limited by using a water cushion and choke with the multisample module. Formations in which seal failures are likely—highly laminated or otherwise heterogeneous formations—or formations that have low permeability can be tested and sampled using the dual-packer module. Instead of a probe and packer to provide a seal, two inflatable packers are used to isolate an interval of about 3 ft [1 m] of formation forming a mini drillstem test. The pumpout module is used to inflate the packers and also to expel mud from between the packers before sampling. Are all the sampling problems solved? Not quite! Resistivity will show transitions only if there is a resistivity contrast between fluids. And, in the presence of more than one phase, the interpretation can be dominated by the continuous phase, making sampling decisions difficult. Also, keeping gas in solution by controlling flowing pressure is possible only if the bubblepoint pressure is known. If bubblepoint pressure is not known, sampling pressure may be too low and the sample could be spoiled by the presence of free gas. A more complete analysis of flowline fluid is therefore needed. Providing a solution for these particular problems is the recently introduced OFA Optical Fluid Analyzer module of the MDT tool. As fluids flow through the MDT‘s flowline, real-time interpretation of the measurements indicates the proportions of oil and water, and gives a qualitative indication of free gas.4 The flowline passes through two independent optical sensors (next page, top). In one cell, absorption spectroscopy is used to detect and analyze liquid, while in the other cell, a special type of optical reflection measurement detects gas. This allows wellsite personnel to decide whether to divert the flow into a sample chamber for retrieval, continue to expel it into the borehole or into a dump chamber, or to increase the sampling pressure above bubblepoint. It has also been used to verify that the formation contains only water or only gas and that a sample is not required. Thus, the sample chambers in the tool are kept available for desired fluids only. Even when oil-base mud is used, it is possible to track the transition from borehole mud, to filtrate, to connate oil as long as the two oils differ in color. After a decision has been made to switch from pumpout to sampling, the OFA module continues to monitor the fluid in the flowline in particular to verify that production remains above bubblepoint. A field test in a vuggy carbonate formation, where conventional probe seals are difficult to obtain, used the OFA module along with several other features of the MDT tool.5 The inflatable dual-packer module isolated the test interval, and the OFA module indicated the fluid in the flowline throughout the test (next page, bottom ). After a pretest to check permeability and seal integrity, 44,000 cm3 [10 gallons] of fluid were pumped through the OFA module using the pumpout module. OFA analysis shows that the fluid changes from borehole mud to filtrate to oil as pumping proceeds. After about an hour, pumpout into the borehole above the packers was stopped and flowline fluid was diverted to the sample chamber for collection. The throttle valve at the chamber end of the flowline maintained pressure above the bubblepoint during sampling. 2. Stewart G and Wittmann M: “Interpretation of the Pressure Response of the Repeat Formation Tester,” paper SPE 8362, presented at the 54th SPE Annual Technical Conference and Exhibition, Las Vegas, Nevada, USA, September 23-26, 1979. 3. Colley N, Hastings A, Ireland T, Joseph J, Reignier P, Richardson S, Traboulay I and Zimmerman T: “The MDT Tool: A Wireline Testing Breakthrough,” Oilfield Review 4, no. 2 (April 1992): 58-65. Badry R, Head E, Morris C and Traboulay I: “New Wireline Formation Tester Techniques and Applications,” Transactions of the 34th SPWLA Annual Logging Symposium, Calgary, Alberta, Canada, June 1316, 1993, paper ZZ. 4. Smits AR, Fincher DV, Nishida K, Mullins OC, Schroeder RJ and Yamate T: “In-Situ Optical Fluid Analysis as an Aid to Wireline Formation Sampling,” paper SPE 26496, presented at the 68th SPE Annual Technical Conference and Exhibition, Houston, Texas, USA, October 3-6, 1993. 5. Badry R et al, reference 3. Oilfield Review Light-emitting diode Gas detector Lamp nThe OFA module with its two sensor systems, one for liquid detection and analysis, and the other for gas detection. Fluid flow Liquid detector Time (sec) Pumpout Module Optical Fluid Analyzer Pumpout Volume Highly Absorbing Fluid 0.0 10000.0 cm3 720 Comments Oil Water Set packer Packer pretest 1440 Start pumpout Pumping filtrate 2160 nOFA log display during sampling using the dualpacker module. Starting at the top of the display, a pretest confirms the seal, followed by pumpout of 44,000 cm3 of fluid from the packerisolated interval. The OFA display— second track from the right—shows flowline fluid change from filtrate to oil. Flow is then diverted to a 1-gallon sample chamber with throttle control to maintain pressure above bubblepoint. Pumping 60% oil 2880 3600 Stop pumpout 44,000 cm3 Start sample Throttling Seal sample January 1994 23 4 Visible Near infrared Crude B Optical density 3 Oil base mud filtrate Crude A Water 2 1 Condensate Diesel 0 500 1000 1500 2000 Wavelength, nm nOil and water spectra. As the wavelength of light increases from the visible to the near infrared, the optical density of water changes from zero—transparent—to a high absorbing peak at 1450 nanometers (nm) and an even higher peak at 2000 nm. The spectra of various oils are also shown, some less optically dense (more transparent) than others in the visible region. A distinctive absorption peak for all hydrocarbons appears at 1700 nm. These peaks are used to distinguish oil from water. Differentiating Fluids The OFA module distinguishes water from oil by differences in optical transmission of light at visible and near infrared wavelengths. The relative intensity of transmitted light—transmittance, defined as the ratio of transmitted light energy to incident light energy—is measured at different wavelengths. Because transmittance of typical formation fluids may vary greatly versus wavelength, it is often convenient to represent optical properties on a logarithmic scale and to use a quantity called optical density.6 The higher the optical density, the less light is transmitted. A plot of optical density versus wavelength is called the absorption spectrum (above ). Three phenomena are primarily responsible for the characteristics of these absorption spectra: molecular vibration absorption, electronic absorption and scattering. Water, as experienced in everyday life, absorbs very little light in the visible region. This continues at the shorter wavelengths in the near infrared region until a resonance in the molecular vibration of the oxygenhydrogen [O-H] bond causes a sudden increase in absorption forming a peak near 1450 nanometers (nm). Another resonance in the O-H bond causes a second, much stronger, peak near 2000 nm. For oils, molecular vibration absorption peaks at 1700 nm, caused by a resonance vibration in the C-H bond. The uniqueness and separation of these peaks permit differentiation of oil and water. 24 Color—the result of electronic absorption—provides one more parameter for liquid identification. Hydrocarbon condensates appear clear or light reddish-yellow, while other crudes may be dark brown or black; undyed diesel and fuel oils tend to be light to dark brown. The color of oil is determined by the shorter wavelengths—blue and green—being absorbed while the longer wavelengths—yellow and red—are allowed to pass. The shorter wavelengths are absorbed by electrons in aromatics and asphaltenes in oil causing a change in their atomic energy state—as the concentration of asphaltenes goes up, more light is absorbed. Analysis of color allows differentiation between oil-base mud filtrate and crudes. Adding a dye to drilling mud could be used to distinguish filtrate water from connate water. If particles mix with fluids—as in drilling mud—or if oil and water emulsify, then light will be scattered. This causes additional reduction in transmitted light. If the scattering contribution to optical density is large, the water and oil peaks may be obscured and difficult to interpret. In this case, the fluid is said to be highly absorbing and is usually interpreted as mud. The transmission measurement is made by directing light from a tungsten halogen lamp to either of two paths. One path is used for calibrating the sensor for minor system drifts before a measurement is taken. The other goes to the flowline, which is coupled to the light via small windows made of sapphire, sapphire being more resistant to abrasion than quartz. Either path can be connected to a spectral distributor where the light is separated into wavelengths—the intensity of each is measured. The wavelengths are chosen to optimize the determination of both the water/hydrocarbon ratio and the color of the fluid in the flowline (below ). Spectral transmission measurements also permit a quantitative holdup analysis in oilwater systems.7 But complications arise when gas is present. Scattering from the interface between liquid and gas, such as by a bubble, attenuates the transmission beam. First, this can add a roughly constant offset to the optical density curves. Second, the presents of gas reduces the height of the hydrocarbon peak, adversely affecting a quantitative holdup analysis. In fact, the hydrocarbon peak height is approximately linearly related to the gas concentration, but relying on this alone does not guarantee that gas is present. A second sensor is therefore used to detect gas and raise a warning flag. Lamp Measure path Source path Sapphire windows Formation fluids Flowline Shutter Spectral distributor Detectors nOFA spectroscopy measurement system. Light from a tungsten halogen lamp is directed along either of two paths. One path is used for downhole calibration, while the other directs the light to the flowline for measurement. Light passes through the flowline fluid via sapphire windows to a light distributor where photodiode detectors, each tuned to a different wavelength, measure the transmission intensity. Liquid analysis is made from these measurements. Oilfield Review Detecting Gas 6. Optical density is defined as the logarithm of one divided by transmittance. In other words, a transmittance of 100% is equal to an optical density of 0, a transmittance of 10% is equal to an optical density of 1 and so on. 7. Water cut is the percentage volume of water produced from a formation over a period of time. Water holdup is a snapshot of the percentage volume of water occupying a pipe or flowline under dynamic conditions. Because of different densities between flowing fluids, there is often a difference in flow velocity—the water holds up the oil. Water holdup will be greater than water cut if the difference in velocity of oil and water is greater than zero. If there is no difference in velocity, then water holdup and water cut will be the same. 8. Light is an electromagnetic wave that oscillates in a plane perpendicular to its direction of travel. Polarized light is light whose oscillations are confined to one plane only. January 1994 Critical angle Approximate band covered by detector array 100 Air Relative reflection intensity, % Gas is detected by measuring reflected polarized light.8 The amount of light reflected at a surface between two media depends on the media, the angle of incidence and whether the light is polarized. The reflection of all light, polarized or not, is governed by the critical angle. At angles of incidence greater than the critical angle, all light is reflected and none transmitted. At less than the critical angle, some light is reflected and some is transmitted. When light is P-polarized—that is, polarized parallel to the plane of incidence—there is a particular angle smaller than the critical angle, called the Brewster angle, that allows 100% transmission. Since values for the Brewster and critical angles differ significantly between gases and liquids, measuring the relative intensity of the reflected light over a range of angles permits positive identification of gas (right ). Using both angles is desirable to detect gas in the presence of liquids. The reflection measurement is made using monochromatic infrared light emitted from a light-emitting diode (bottom, right ). This is polarized and passed through a cylindrical lens, prism and sapphire window into the flowline. An array of six detectors measures the intensity of light reflected from the sapphire/fluid interface at discrete angles, from just below the Brewster angle for air, to just below the critical angle for water. Calibrations are carried out with air and then with water in the flowline. A gas flag is currently generated by a simple algorithm that uses the shape and location of the measured reflectivity curve between the 100-percentair and 100-percent-water curves. Reflection responds to the interface between two materials, in contrast to transmission spectroscopy, which samples the bulk material. Hence, the gas detector responds to free gas at the surface of the sapphire window, which may not necessarily represent the Water Oil 95 90 10 5 0 0 15 45 60 Angle of incidence Less than critical angle Lens and polarizer nRelative reflection intensity of P-polarized light for air, water and a typical oil. The relative reflection intensity is shown for varying angles of incidence for air, water and oil. The reflection intensity decreases to zero for P-polarized light at an angle of incidence called the Brewster angle. As the angle of incidence is increased, the reflection intensity increases until total internal reflection occurs at the critical angle. Brewster angles and critical angles are different for air, water and oil. Measuring reflection intensities over a range of angles provides the basis for the gas detection sensor. Brewster angle Sapphire prism Detector array LED Formation fluids Flowline nGas detector optics. Light from a light-emitting diode (LED) is polarized and spread over a range of incident angles through a sapphire window at the flowline. An array of detectors measures the reflection intensity over angles from just below the Brewster angle for air to just below the critical angle for water. Comparing intensities with those obtained from master calibrations allows gas detection flags to be raised on the log. 25 volume fraction of gas present in the flow. However, liquid films on the sapphire window do not impede gas detection. Distinguishing between single phases of oil, water or gas in the flowline using the OFA module is straightforward: the gas detector shows either gas or liquid, while the absorption spectroscopy peaks define the liquid. If the flow is two or three phase, a more complex analysis is required to quantify volume fractions. This is possible for liquid holdup as the absorption spectrometer measures optical densities across the flow. However, the gas detector measures light reflected at the liquid sapphire interface, which generally produces a larger signal for greater gas fractions, but may not be representative of gas holdup. At present, water holdup is calculated directly from the calibrated responses of the detectors tuned to the two water peaks relative to those tuned to wavelengths at which water has very little absorbance. Subtraction from unity then gives hydrocarbon holdup. In addition, the log presents an oil indicator by shading the separation between the output of the detector tuned to the oil peak and the output of the detector tuned to a wavelength between the oil peak and the 1450nm water peak. The magnitude of the separation gives an approximate measure of the oil volume fraction in a gas-oil mixture. When combined with the hydrocarbon holdup estimate, this can yield a rough appraisal of oil holdup. Since scattering reduces light transmission, a shift from the baseline of the optical density curves shows the magnitude of scattering. Independently, the gas detector flag indicates the presence of gas. The recording also presents optical densities measured by each transmission detector. These are tracked in time, and give additional indications of changes in flowline fluid composition. For a more formal approach to future interpretation development, see “Effective Flow Stream Model,” (below ). Examples of Sampling Using the OFA Module In a field test in a naturally fractured formation, the dual-packer module was used to provide a good seal and the OFA module was used to monitor the flowline fluid composition during a pretest and subsequent sampling operation. 9 A display of each channel from the OFA spectroscopy measurement system shows the optical density measurement recorded throughout (next page ). Channels 1 to 6 sample the spectrum below the first water peak and respond to color. Channels 6 to 10—channel 6 is repeated—are the water/oil channels and respond to changes in the region of high 9. Smits AR et al, reference 4. Effective Flow Stream Model Flow regime Flow direction Model Light Oil Water Oil Water nTransforming flowline contents to effective flow stream (EFS) model. The two-phase flow stream cross sec- tion (left) is grouped into different flow regimes and simplified (right) into a model consisting of a number of optical components in parallel and series with the light. The calculation of oil and water holdup and estimations of fluid coloration and light scattering involves the solution of 10—one for each wavelength—simultaneous equations given by this model. A somewhat more formal approach to interpreting The EFS model provides a method for reducing the as occurs from particulates, emulsions and gas- complex flow regimes along the flowline through dynamic flow to a static optical equivalent (above). liquid interfaces, also contributes to attenuation by the OFA absorption sensor uses the Effective Flow The transition from flow stream to static model is removing photons from the optical beam path. Stream (EFS) model. It employs a set of transmis- made by grouping slug flow, emulsion or bubble This is treated for both mixed and segregated sion equations, one for each of the detectors. The flow, and layered flow, so as to segregate the phases as a wavelength-independent equivalent unknowns correspond to oil and water holdups, transmissions accordingly. The EFS model reduces absorption in series with that of the flow compo- degree of oil coloration and amount of scattering. this to a simple partitioning of series and parallel nents. The EFS model is also the basis for the transmissions. This model represents the mea- ongoing development of additional, improved sured transmission as just the time average of interpretation algorithms. transmissions through oil, through water, and through oil and water in series. Scattering, such 26 Oilfield Review Pumpout Motor Time, sec Oil Indicator SC 1 Valve Pos’n Highly Absorbing Fluid Pumped Volume 0 cm3 Water Fraction 20,000 Hydrocarbon Fraction Pressure 0 psia 5000 Gas 1 0 0 Optical Density Color Channels 1 Comments Water, Oil Channels 150 Inflate packer Packer pretest Start pumpout 1050 Pumping filtrate 1950 Pumping oil Stop pumpout 2850 Start sample Throttling Change throttle 3750 January 1994 nFormation sampling with dualpacker and pumpout modules. The top of the log (0 to 350 sec) shows the pumpout module being used to inflate the packers of the dual-packer module—the OFA reading shows mud. During the packer pretest (400 to 750 sec) the OFA reading again shows mud. Fluid is then pumped out from between the packers changing from mud to filtrate (showing as water), and finally, at 2200 sec, to oil. Pumpout is stopped at 2550 sec and the sample chamber opened. During sampling, throttling controls the flowing pressure to ensure gas stays in solution. The 10 optical density channels are displayed to give a visual presentation of density changes throughout the test. Channels 1 to 6 sample the fluid spectrum below the water peak at 1450 nm and give an indication of changes in color. Channels 6 to 10 are the oil/water channels responding to changes above 1450 nm. Seal sample 27 10. Smits AR et al, reference 4. 28 Time, sec absorption peaks. The character of the spectrum at any time can be visualized by sweeping the eye from left to right across the log and noting changes in optical density. Hydrocarbon and water holdup curves are shown in the adjacent track. In this example, the absence of a gas flag suggests that the hydrocarbon is all oil. Confirmation of this is visible on the oil indicator, which correlates well with the estimated hydrocarbon fraction and with optical density in the color channels. An initial flow of mud is visible on the optical density curves. These produce the highly absorbing fluid indicator that appears at the beginning of the log. The highly absorbing fluid flag appears whenever all of the detectors sense essentially no light. The OFA response to gas is shown in another field test where alternating slugs of water and gas occur (right ).10 The optical density track shows no absorption in the color channels, suggesting absence of oil, and increased density in channels that respond to the water absorption peaks. Water- and hydrocarbon-fraction curves step between 0 and 1 as slugs pass, and the gas flag responds similarly. The oil indicator confirms that the hydrocarbon is gas, and not a light oil that is transparent in the visible region—no shading can be seen. The transmission differences between gas and water slugs cause oscillations in the optical density curves—and, hence, the oil indicator curves—and not because of changes in scattering. The very slight time shift between Oil Indicator Pressure 0 psia 5000 Gas 1 Optical Density Hydrocarbon Fraction Water Fraction 0 0 1 Color Channels Water, Oil Channels 5520 5580 nSection of log display during acquisition of a gas sample. The OFA spectroscopy measurement alternates between water and hydrocarbon. The color channels show transparent fluid, whereas the water/hydrocarbon channels oscillate. The gas detector shows oscillations between liquid and gas. The conclusion is that alternating slugs of gas and water are moving along the flowline. the optical density and gas-indicator curves results from the small flowline distance between the two sensors. Comparing measurements by the OFA module with surface observations on the chamber contents must be done with care—gas can come out of solution and liquids can vaporize. Preserving downhole temperature and pressure is necessary for valid comparisons since even returning them to their downhole values may not guarantee recombination of phases to their original condition. So far, OFA outputs have mostly been consistent with other knowledge about fluid production in individual wells. Several field tests have shown good agreement between OFA measurements and chamber analyses for oil and water, even though the OFA module measures holdup and the chamber contents represent timeintegrated cuts. Other tests show quantitative disagreements that could be caused by several factors such as not flowing through the OFA module during sampling or sampling at substantially different flow rates to those during pumpout. Further developments in OFA measurement and interpretation are in progress and new applications can be envisioned: aiding interpretation of drawdown pressure tests, providing new information on formation invasion and making quantitative estimates of bubblepoint under downhole conditions. —JT, AM Oilfield Review