_RP_11L-1988_scan.pdf

ERRATA October 1,1988 to API RP llL RECOMMENDED PRACTICE for DESIGN CALCULATIONS for SUCKER ROD PUMPING SYSTEMS(Conventional Units)FOURTH EDITION JUNE 1,1988 Issued by AMERICAN PETROLEUM INSTITUTE Production Department 2535 One Main Place Dallas TX 75202 NOTE:Paragraphs AIO through A17 1cere inad1ertently left out of the fourth edition of APIRP J lL.These missing paragraphs are included in this Errata.AlO.In determining the natural frequency,thevelocity of force propagation a plays a key role.The theoretical value for this velocity is about17,000 feet per second.In practice,however,it hasbeen found that lower frequencies normally occur.The speed of sound in long narrow rods is usuallysomewhat lower than in normal size vessels.Also,the effect of the rod coupling can cause an apparentincrease in density resulting in a decrease in propagation velocity.In practice,it has been found thata equals about 16,300 feet per second.With thisvalue,the nondimensional pumping speed can becalculated by:N NLNo=245,000 Fo All.Another dimensionless parameter of importance m describing the behavior of the sucker rodstring is the dimensionless rod stretch,Fo/Skr,Inthis parameter,the spring constant,k,is involved,For untapered rod strings,it is evaluated by therelation:EA lcr=For tapered rod strings,the spring constant is computed from the familiar reciprocal formula:1 1 L1 L27 =E A1+A2+.JThe complete term,Fo/Skr gives the rod stretchcaused by static application of the fluid load as a percentage of the polished rod stroke.For example,=0.1 means that the rod stretch is 10%of the polishedrod stroke when the fluid load is statically applied.At very low speeds where static conditions areapproached,the dimensionless rod stretch and pump stroke are related as follows:Sp 1 _ FoS-Skr At higher speeds,this relation breaks down as dynamic effects become more important.A.12.The charts are used to determine the dependent parameters as defined in the report.Threeforces of particular importance are determined fromthe charts.F1 is the fluid load plus the maximumdynamic effect on the up stroke.F2 is the dynamiceffects during the down stroke taken away from theload on the polished rod.Fs is a force which will givehorsepower when applied to the full stroke lengthat the speed of the pumping unit.Al3,The surface dynamometer card which isgenerated in the analog computer is independent ofthe weight of the rod string involved.The shape ofthe card will be the same and a different rod loadwill simply shift the card up or down in relation tothe zero line.In calculating all parameters excepttorque,this can be handled very easily by calculatingfluid and dynamic loads independent of the rod load and then adding in the rod load.In the case of torque,the amount of rod load is important in determining the torque involved in the unit.All torquevalues were calculated using a rod load of:Wrt Sk,.=0.3If a rod load different from this is used,a correctionmust be made as shown in Fig.4.6.AU.In the analog model,it has been assumedthat the tubing is anchored and no tubing motion occurs during the pumping stroke.If the tubing isunanchored,then a correction must be made for the shortening of pump stroke which will occur.This happens because the tubing shortens when theload is transferred to the rod string and thenlengthens during down stroke when the fluid loadis transferred back to the tubing.A correction inpump stroke is made by simply calculating theamount of tubing stretch which will occur with the fluid load used in the design and subtracting thisamount of stroke from the net plunger stroke atthe bottom of the hole.This should give a reasonablecorrection.It will not be absolutely correct due tothe fact that dynamic effects occur in the tubingstring as well as in the rod string and these dynamiceffects are not taken into account.A15.An average conventional pumping unitgeometry has been used in the simulation.This isa conventional unit with the counterweights in phasewith the crank and the tail bearing being over theslow speed shaft at midpoint of the stroke.For unitswith drastically different geometry from that assumed,the simulation will not be accurate and thevalues calculated will be more approximate.It isbelieved,however,that the values for maximum andminimum loads and for pump stroke will be reasonably good.The calculated value for torque will notbe even approximately right and a torque calculation must be made by some other method.A.16.Pumping unit motion was assumed based ona medium slip motor.Generally speaking,a higherslip prime mover results in slight decreases in themaximum load and a slight increase in the minimumload but also tends to reduce the subsurface pumpstroke;therefore,some error may be introduced inthe calculations if a prime mover with a considerably different slip characteristic than a medium slipelectric motor is used.Also,the assumption is madethat no friction occurs in the stuffing box or in thepump itself.This,of course,is an unreal assump