US2376542A - Method of removing samples from flowing streams - Google Patents

Description

y s. o. KIMMEILL, 2,376,542

METHOD OF REMOVING SAMPLES FRDM FLOWIHG STREAMS,"

; "Original Filed Jufy 27, 1940 4 V Sheets-Sheet." J.

flaw/79hr fiecaralhg 01-292"; Me fer Jepamfqf ATTC'DRNEY May 22, 1945. I 2,376,542

ua'ra'on OF REIVIOVING' SAMPLES FROM FLOYIING .swmams I G- O. KIMMELL Original Filed July 27, 1940 4S1'ieets-Sheet 2 Illa L III R In WI] fill Ill 7 who May 22 1945 o. KIAMMELL METHOD OF REMOVING SAMPLES FROM FLOWING STREAMS Original Filed July 27-, 1940 4 Sheets-Shae? 3 V// 0 I// n// A/// I May 22,1945. G. o. kIMMELL METHOD OF REMOVING SAMPLES FROM FLOWING STREAMS Original Filed July 27, 1940 4 Sheets-Sheet 4 INVENTQR ATTORNEY Suing/10 L I V\\\ 'd' an: we?

U I l/ll/l/l/Il/l/J/lll/I/IIWIIWIIr/l/fl/l/I/Jl/IlId NM Patented May 22, 1945 FEOWING Gal-man 0. ii o o I: W-Eity, QOlsla. @riginal application July 27, 940, Serial No. 3%,063. mvided and application April 3, 1941, Serial No. 386,679

s @llaims. (s11. 33-422) This invention relates to a method ofremovrelation to i better illustrate the construction mg samples from flowing streams, particularly thereof.

turbuiently flowing composite streams Sllchfas Fig. 9 is a detail perspective view, partly in v the flow from high pressure distillate wells, the longitu s at one o the e i present application. being a'division of my co- 3 DQ IB pending application Serial No. 348,063 filed July islll is a s i view ofja smellefnezzle- 2'7, 1940. Wells of this character have optimum Fig. ii is a cross-section through the spider conditions of operation under which the maxi-- Oh the line of um mount'of liquid hydrocarbons may be r Mg. 12 is a detail sectional view through the covered it the well pressures are properly mainm tubing ehdlsampling hezzler Showing flow tained. The required pressure varies with the cheleetemties through the tubing when the saturation of the reservoir gas with the recover- W100"? in the heme is greater than the able constituents. Consequently, it is essential to veleeity through the tubingobtain accurate samples of the flow at selected 3 3 is 81111.11 Viewshowmg the pressures to determine the reservoir characterls- 8 characteristics When the Velocity in the tubing tics and the most desirableabove ground condiis greater then the flew Velocity through the tlons for therecovery of the maximum amount of malellquid hydrocarbons. A vit l factor in making Fi 14 is a similar-view when the velocities are an analysis of a .well is that of obtaining a true M equalrepresentative sample oi the well flow at well 15 is e eph depicting deteebteihed head pressures. ins tests of the flow in atrflactual weilrl- It is, therefore, the principal object of the pres- Referring more in de to h ewmgsi .ent invention to provide a method of obtaining I ates head 0! well drilled e re resentative sam les of th 11 t t f 1 gas producing hmhahoh and equipped with W511 head Dressure during fi w lhhe e o h a m a well tubing 2 which extends thmhgh. the casing In accomplishing this and other objects oi-the" a and toms conductor through when the wen invention, as hereinafter pointed out, I have provided improved details of structure, the preferred form of which is illustrated in the accom- 30 seal the space about the upper end of the casing 3 and tubing 2. The upper end of the tubing'is flows into a separator -l. Pressure is maintained within the well by means of a casing head 5 to pahyihg drawings wherein provided with a cross-fitting 6 having lateral Fig. 1 is a diagrammatic view of the apparatus used in sampling a well stream discharged from a producing formation under natural formation pressures or through one of the repressuring as methods well known in oil field practice.-

Fig. 2 is an enlarged longitudinal section through the jack tor-introducing the sampling branches I, at least one of which is connected by a pipe 8 with the-liquid and gas separator 4 under control of a valve 9. p

The separator I is of any conventional design and has a gas outlet connection? ill with a discharge line H equipped with a recording orifice meter l2, whereby rate of flow of the gas coihnozzle into the well tubing against the well head ponent of the wen stream is recorded The lower pressure.

portion or the separator is equipped with the I Fig. 3 is an enlarged cross-section on the line usual quid discharge connection 3 having 3-4 of Fig. 2.

Fig. 4 is a detail perspective view, partly in 3 3 23 32? 6 flow Hue leading to a storage ti n. of ne of the tu e ripp P The terms"distlllate wen or "condensate well" 5 is crass-section m the line are nowin good engineeringand legal use to de- Fig. 2. it v ribe a gas well which yields an essentially colorll'ig. 6 is a cross-section on the line 8-8 of less product under m l iti n of o ra- Fl 2. z tlon. For this type of well the flow comprises a Fig. 7 is an enlarged section through the uP- composite stream of hydrocarbon liquid and gas. Der P r i n of th W Showing placem of 5 The present conditions of'supply and demand are he mpling nozzle and the spider for re inin f responsible for the inability of the producer to the nozzle in substantially coaxial position withsell the gas phase from his distillate we'll. Under in the welltubins. proration rulings, it the producer cannot sell the Figija is a'detail perspective view of the parts gas, he is not allowed to waste it for the recovery oi the spider clamp shown in disassembled spaced oi the distillate alone. The growing practice is to haudlelarge flows from the well, extracting the distillate therefrom, and pumping the unliquefiable components back into the reservoir or producing formation. .This practice eliminates waste of saleable gas and allows the extraction of the I valuable distillate. Therefore it is extremely dehead pressure for use in the small scale unit In the small scale unit pressures can be varied from full flowing well head pressure down through the retrograde range.

- I have, therefore, provided for taking samples simultaneously'with the flow. This is effected by' introducing a sampling tube Ii into the well tubing so that a proportionate part of the flow passes through the sample tubing to a test separator l1 having capacity for determining the amountof recoverable liquid hydrocarbons at al pressures up to the pressure of the well.

In order that the tube Iii may be introduced into the well tubing under high well pressures, the tubing below the cross-fitting 6 is provided with one or more gate valves l8 to close on the well flow through the tubing so that the axial aligning branch IQ of the cross-fitting may be unplugged and equipped with a fitting having a bore 2| adapted to receive the reduced neck 22 of a jack-screw 23. The jack-screw 23 has a polygonal-shaped portion 24 adapted to be engaged by a wrench, to turn the neck 22 into the threaded opening 2|. Extending upwardly from the portion 24 is a cylindrical shaft-like body 25, having a thread 28 formed thereon fora purpose later described. The iack-screw-is further provided with an axial bore 21 continuing through the threaded neck 22 and of a size to slidably pass the sampling tube l6, as shown in Fig. 2. The upper end of the jack-screw is" counterbored, as at 28, to receive a packing element 29 adapted to sealabout'the tubing. 'The packing 29 is retamed in sealing position by means of afollower 30, having a threaded portion 3|v engaged with an internal bore .32 formed in a reduced upward extension 33 of the jack-shaft, the upward extension being provided with slots 34 whereby the extension is adapted to be contracted about the threaded body 3| by means of a clamping 'ring 35. The clamping ring 35 is normally of a size to be freely mounted on the reduced extension and has laterally extending ears 36 through which a fastening device, such as abolt 31, is extended to tighten the ring about the extension and prevent loosening of the packing follower 30 when desired. The follower 30 also has a polygonalshaped flange 38, provided with radial sockets 39 to receive a spanner wrench or a turning bar (not shown), whereby the follower is moved to andfrom compressing relation with respect to the.

packing. The follower further includes an uppe threaded portion 40, carrying a sleeve 4i having a plurality of laws 42 adapted to grip the sampling tube when tapered portion 420i the laws are engaged in the tapered bore 44 of a clamping causing the slip Jaws to grippingly engage the sampling tube.

The collar 45 is retained between thrust bearings 46 and 41 carried within a, bearing housing 48, having an internally threaded portion 4! engaging the upper threaded extension 40 of the packing follower. The bearing housing 48 has an annular portion 50 provided with radial sockets 5| to receive a suitable actuating bar, whereby the bearing housing may be rotated relatively to the packing follower to cause the collar 45 to move downwardly in contacting relation with respect to the slip jaws 42 to cause the slip jaws to grip the sampling .tube when the tube is passed therethrough, as later described.

Rotatable on the threaded jack-screw is a nut 52 fixed within the lower end of a cage-like housing 52 which extends upwardly over the slips 42 and carries a chuck 54. v The nut 52 has radial $00kefl5 55 registering with openings 56 in the wall of the housing 53 to accommodate handles 51 whereby. the housing may be rotated upon the jack-screw. The housing 53 is of such length to accommodate the bearing housing 48 when the nut is at the lower'end of the jack-screw, so that a cap plate 58 'forming the upper end of the housing amply clears the upper end of the bearing housing 48. g

The upper chuck includes a body 59 having a threaded neck 60 mounted within a threaded opening 6| of the plate 58. The body 59 also includes an upper threaded periphery 62, carry ing a bearing housing 63. Projecting upwardly from the threaded portion 62 is a reduced extension-84 having hooked jaws or slips 6!. The

jaws are tapered, as at 66, to receive the tapered bore 81 01a contracting collar 68, similar to the collar 45, previously described.' The collar 68 is retained within the bearing housing 63 between thrust bearings 69 and II. The upper bearing housing alsoihas a portion. II provided with radially extending handles 12, whereby the bearing housing may be rotated to move the collar 68 to and from contracting relation with respect to the Jaws or slips.

The tube I6 is of sufllcient length so that when introduced into the tubing the terminal end 12 thereof is located at a point within the well tubing wherein the flow is not influenced by eddy currents or other disturbing influence. I

I have found that the wall thickness of the tubing at the inlet end thereof will form an obstruction and prevent taking of accurate samples of the well fluid. However, I have solved this difflculty through use of a nozzle I4 having a thin body portion 15, the inner surface of which is a perfect cylinder and the outer surface a slightly tapering cone to form an extremely thin entering edge 16 projecting within the well stream, the edge being of such thickness as to avoid eddy currents and resistance which interfere with taking of representative samples when I other conditions are provided for, as later describeda In order to' mount the nozzle upon the end .of the sampling tube, the sampling tube is preferablyprovided with a reduced threaded neck 11 forming a seating shoulder 18 whereon a socket 19 of the nozzle 14 is threaded as shown in Figs. 12 to .14 inclusive. The interior bore of the nozzle at the edge I6 is determined with extreme accuracy so that the effective area thereof is collar 45 that is sleeved thereover, as best shown in Fig. 2: It is thus obvious that the samplin readily proportional with the effective crosssectional area of the well tubing.

that the velocity .sampling nozzle does not touch or nearly touch the wall of the tubing. Therefore I provide the lower end of the sample tube with a centering spider 80 that is slidable thereon to an adjusted position with respect to the inlet of the nozzle so that the spider is located a sufficient distance above the nozzle to avoid interference with taking of an accurate sample.

In the illustrated instance, the spider includes zle will be taken from a hypothetical sampling cylinder coinciding with the hypothetical sampling cone and there is no tendency for a part of the denser phase to by-pass or enter a part of the nozzle tip, and a representative sample is removed.

I, therefore, control velocity through the sam- I pling tube, as now to b described. The test sepa collar 8| slidable on the sampling tube and connected with bow-shaped springs 82 having their upper ends connected to a similar collar 83, the collar 83 having a threaded neck portion 84 terminating in slips 95 having tapered peripheral faces engaged within the tapered bore of a nut db, the nut 86 being freely rotatable about the sampling tube and having a threaded portion it! to engage the threaded portion 84 of the upper collar 83. Thus when the nut is tightened about the slip portion 85, the upper collar 83 is securely gripped in an adjusted position with respect to the sampling tube it, to retain the inletofthe nozzle'in substantial alignment with the axis of the tubing.

' i have found that anotheressential requirement in obtaining a representative sample, is through the thin body portion it of the nozzle 74 be maintained identical with velocity through the well tubing 2. This is readily explained by observing Figs. 12 to ,ld inclusive. In Fig. 12, the velocity V2 in the nozzle tip is greater than the velocity VI in the sampling section. Consequently, more material. will be drawn into the nozzle tip than would normally pass the section occupied thereby. The extra material entering the nozzle would be taken from a zone between a hypothetical sampling cylinder b in a hypothetical sampling cone a. A particle of liquid in the denser tween the hypothetical sampling cone and the hypothetical sampling cylinder, because of its inertia, would tend tobreak through the hypothetical cone a and by-pass the nozzle as shown by the small arrow. This would cause a loss of the denser phase in the sample and the result would be the recovery of a. sample in which the ratio of theless dense. to the dense phase would be too high.

In Fig-13, the velocity V3 at the nozzle tip is lessathan the velocity VI in the sampling section of the well tubing. In this instance, less material will be drawn-into the nozzle tip 3 than would normally pass the section occupied by the nozzle tip. The material entering the nozzle would be taken from a zone between a hypothetical sampling cylinder 0 and a hypothetical cone d. A particle of the denser phase in the zone between the hypothetical cone and cylinder, because of its inertia, would tend to break between the hypothetical sampling cone and enter the nozzle 8. As shown by the arrow indicat-- ing the path of the denser particle, this would cause a gain of the denser-phase in the sample and the result would be the vrecovery of a sample in which the ratio of the less dense denser phasewould'be to low.

In Fig. 14, the velocity VI in the nozzle tip is identical to the vel0city'Vl in the sampling section of the well tubing. The same amount of material will be drawn into the nozzle tip as would phase in the zone bephase to the arator H is connected with the tubing It by means of a pipe 87. Associated with the pipe t'l is a preheater 88 and a stabilizer 89 located on the respective sides of a control valve till. The preheater includes a coil 9| through which the sample of the well fluid is circulated, the coil being contained within a housing 92 through which steam or .other heating medium is circulated. The stabilizer includes a similar coil 93 contained within a housing 94 through which a cooling medium, such as water, is circulated. The temperature of the sample may thus be regulated by controlling flow of the heating and cooling medium test separator. The gas content of the sample is discharged from the separator through a pipe 95 having connection with a coil 96 of an afterheater 98, similar in construction to the preheater 92, previously described. From the coil it the gas is discharged through a control valve tit into a critical flow meter Mil, including a chamber it! equipped with. a pressure gauge tilt and a thermometer W3, whereby pressure and temperature within the chamber may be noted and brought to the required values. The chamber Mi has athreaded neck int projecting from the discharge end thereof which formsa seat for an orifice disk tilt, the disk Hi5 being retained by a cap M6 threaded on the neck mt. The orifice Hill of the disk I05 has a size proportionate to the inlet of the nozzle, whereby the rate of flow may be maintained so that the velocity through the nozzle is equal to the velocity ofthe well flow passing the exterior of the nozzle. The separated liquid is taken from the test separator I? through a pipe I09 under control of avalve ltd.

In practicing my method, the well flow is shut f off to the separator by closing the gate veg/es i 3.

The jack is then applied by screwing theth eaded neck 22 thereof within the threaded bore 2! of the fitting 20, the tube It being contained withthrough so that the nozzle as just above the uppermost valve W. The jack-nut 52 is then moved to the upper end of the jack-screw 25. This is effected by rotating the cage-like housing 53 by means of the handles 51. The chuck M is then tightened about the sample tube I6 so that the slips thereof grip the tubing. The lower set of slips, however, are left in disengaged condition. Setting of the upper slips is effected by rotating the bearing housing by manipulation of the bandles l2. After the discharge end of a sample tube 16 has been appropriately connected with the preheater andother test equipment, the gate valves I8 are opened so that the pressure acts dinormally passthe nozzle section occupied by the nozzle tip, because the material entering the nozrectly against the nozzle of the sampling tube, but the slips of the upper chuck hold the'sampling tube in position against pressure of the flowin fluid 'which'is now being discharged through the pipe 8 into the separator. The jack-nut '52 is then rotated to the bottom of the jack-screw, causing the slips carried therewith to. move the nozzle further into the wen tubing. when the nutis in ried further into the tubing.

its lowermost position, the slips or the lower chuck-are tightened about the sampling tube and end of'the screw, the slips of the upper chuck are set in gripping relation with the sampling tube, after which the slips of the lower chuck are .loosened so that when the nut 52 is rotated and moved in a downward direction, the nozzle is car- This process is continued until the nozzle is in desired position. The well may flow in heads upon initial opening of the valve 8, and no attempt is made to take a sample until after the well is flowing steadily. When the well is flowing steadily, the pressure orifice meter l2 and a'proper orifice plate Iii! is selected and applied to the critical flow meter so that the flow through the sampling tube to the test separator approximately equals the veloci y flow exteriorly oi'the nozzle.

The sample data obtained during actual tests of a well are repres nted in the graph in Fig. 15. A inch nozzle as examined for itsabllity' to remove a representative sample at two rates of flow from the well; six and eight million standard cubic feet of separated gas per day. 'A V; inch nozzle was examined at a rate of flow of four million cubi feet of separated gas per day.-

As shown-in the rt, three samples were taken at a four million ate using inch nozzle; one when the velocity ratio (nozzle velocity to sampling section velocity) wasadiusted to substantially one, another at about'.82, and a third at about .47. Samples were also taken with a 1?! inch nozzle at a six million rate with a velocity ratio adjusted to about 1.2 1,' .78, and .65. Additional samples'were taken with a 5": inch nozzle at an eight million rate. The liquid content of all the respective samples per cubic foot of gas is shown plotted against velocity ratio. An averagea curve is shown giving each point approximately equal value which curve is within the range of experimental error. Under all three conditions. a ratio of liquid to gas of approximately 17.7 cubic centimeters per cubic foot was obtained in the experimental unit when the ratio of velocities inside and outside the sampling nozzles were equal, but any sampling condition resulting in a deviation of unity ratio oi velocities inside and outside the nozzle tip gave an erroneous indication of the liquid gas ratio, as shown by the chart.

With a truly representative fractional part of the well stream available it is possible to examine the characteristics of a total well stream in a small scale unit. Because of increased acand flow velocity are obtained from the recording curacy and convenience of operation 0! a small scale testing unit, it is possible to'take data with a greatly increased rate oi speed and at the same time materially improve on accuracy obtained in full scale tests, all this being possible through my improved method 01 taking truly representative samples, as above described.

-What I claim and desire to secure by Letters Patent is: I

1. The method of sampling a distillate well stream containing both liquid. and gas phases including, concurrently removing a predetermined cross-sectional portion of said stream completely within the confine of said stream, maintaining a rate of flow'velocity within the cross-sectional portion substantially equal to the velocity of the stream surrounding said cross-sectional portion atthe point of removal, changing velocity of the sample I stream, taking separate samples after each velocity change, and comparing the amount of liquid hydrocarbon in the respective samples with change in velocity ratio for preparing data to indicate the liquid content at equal velocities inside and outside said predetermined cross sectional area. I

2; The method of sampling a distillate well stream containing both liquid and gas phases including, flowing the well until the flow is steady.

concurrently removing a predetermined crosssectional portion of said stream at said well head pressure and completely within the confine of said stream, maintaining a rate of flow velocity within the cross-sectional portion substantially equal to the velocity of the stream surroundin said cross-sectional portion at the point of removal, changing velocity of the sample stream,

taking separate samples after each velocity change, and comparing the amount of liquid hydrocarbon in the respective samples with change in velocity ratio for preparing data to indicate the liquid content at equal velocities inside and outside said predetermined cross sectional area.

3. The method of sampling a distillate well stream containing both liquid and gas phases including, concurrently removing a predetermined cross-sectional portion of said stream substantially coaxial of said stream, maintaining a rate of flow velocity within the cross-sectional por- -tion substantially equal to-the-velocity of the stream surrounding said cross-sectional portion,

changing velocity of the sample stream, taking separate samples after each velocity change, and comparing the amount of liquid hydrocarbon in the respective samples with change in velocity ratio for preparing data to indicate the liquid content at equalyelocities inside and outside, said predetermined cross sectional area.

GARMAN ,0. 'KIMMELL. I

Images