CN101415905A - Method for optimising the production of a cluster of wells - Google Patents

Abstract

The present invention relates to a method to optimise production of a cluster of wells on the basis of an estimation of the contributions of individual wells to the production of the cluster of wells, tailored to the particular constraints and requirements of the oil and gas production environment. The wells in the cluster may differ in terms of nature and flux of its effluents, and/or mode of operation, stimulation and/or manipulation. The wells may also produce from multiple subsurface zones or branches. The wellheads of the wells in the cluster may be located on land or offshore, above the surface of the sea or on the seabed. The method according to the invention may be used to generate one or more optimisation models, taking into account only significantly relevant well and production syatem characeristics and effects.

Description

Optimize the method for well group output

Technical field

The present invention relates to a kind of method that is used to optimize the output of hydrocarbon production system, described hydrocarbon production system comprises hydrocarbon production well group and relevant fluid separating assembly.

Background technology

Usually, the fluid stream of being produced by each well in the well group is mixed into multiphase flow and sends to the fluid output tube road via fluid separating assembly (comprising one or more volumes and/or production separator (bulk or production separator)) in one or more headers (manifold) pipeline, nominally be used to carry and the stream that separates at least of selling oil, gas and/or other fluids.

The problem relevant with the fluid flow pipe reason in the exit of volume or production separator is that this fluid flows and comes from the mixing flux of the some or all of wells in the well group, and can not provide about being made up of and the information of flux the fluid of single well production.Thereby, usually can not be real-time or the independent flux of the fluid produced by single well of instantaneous ground accurate tracking.Because it is variable in time to follow the tracks of single well yield and well yield and performance, will directly cause when different wells are regulated, fully ground well characterizes so that predicted flow rate.In addition, well yield is because header and eliminator handle whole potential productions limited in one's ability of artesian well and frequent interaction.For example, the gas output in a well is too high may reduce oil yield total in the well group.

Another relevant problem of output with monitoring and control hydrocarbon production well is that well may produce the irregular slug flow that may comprise crude oil, gas, water and/or condensate in the mixture of crude oil, gas, water and/or condensate and the output.Heterogeneous quantifier is usually too expensive, limited seriously that opereating specification and the too complicated flow duct (flowline) that can not be installed in single well go up so that the gentle single component of oil, water that allows real-time continuous ground measuring well to produce, especially when the life period generation significant change of the multiphase flow dynamic characteristic of well at well.Secondly, these heterogeneous quantifiers are when starting and/or also need calibration sometimes.Thereby at the utmost point in most cases, the fluid-withdrawal rate of single well usually can not be by measurement directly, accurately, continuously or in real time.

International Patent Application WO 03/046485 discloses output metering and well surveying test system, wherein the cumulative production in whole field with "nine squares" is in the measured downstream of production separator, in this production separator, isolate the production flow point (fraction) of crude oil, water, natural gas, solid and/or condensate, and can accurately monitor the crude oil of production and/or the flux and the composition of other flow points.The accurate measurement of this well cumulative production to whole field with "nine squares" with carry out simultaneously and contrast measuring of well effluent flow measurement upstream that each single Jing Chu carries out simultaneously than inaccuracy.

International Patent Application PCT/EP2005/055680 that the applicant submitted on November 1st, 2005, " be used for determining the method and system of single well " to the contribution of well group output, described a kind of method and system, hereinafter it is called as " the overall monitoring in real time of output " (PU RTM).PU RTM method allows accurately to estimate in real time the contribution of single well to total mixing output of the well group of crude oil, gas and/or other fluid production wells, and it is based on the well model of being derived and used the dynamic data regular update of mixing output by well surveying examination data.PU RTM method also need not monitored Jing Chu configuration multi-phase flowmeter at each.

Summary of the invention

An object of the present invention is to provide a kind of method and system of the optimization well group output in the special constraints of estimating to adapt on single well is to the basis of the contribution of well group output the production of hydrocarbons environment and requirement.Well in the well group may be different aspect the character of its product and flux and/or operator scheme, excitation and/or manipulation.Well also may be produced from a plurality of underground payzones or branch.The well head of the well in the well group may be positioned at land or sea, on the sea level or in the seabed.The method according to this invention can be used to generate one or more optimization models, only considers the relevant especially well and the feature and the effect of production system.

According to the present invention, a kind of method that is used to optimize well group output is provided, the well effluent of described well group stream mixed and fluid separating assembly be separated into crude oil, gas and/or other fluids to the small part separated flow, described method comprises:

A) each well is carried out well surveying examination, during this period, tested well yield be changed and the manufacturing variables of one or more single wells monitored;

B) according to the data that obtain by well surveying examination obtain with by by the estimation model of each relevant well of the variation of the well manufacturing variables of the variation of the flow pattern of the product of well logging production and monitoring;

C) make normal miscella of well input and/or gas production;

D) during step c), the dynamic fluid flow pattern to the small part separated flow of flow meter monitoring crude oil, gas and/or other fluids to the small part separated flow of crude oil, gas and/or other fluids by being arranged in the fluid separating assembly downstream;

E) during step c), monitor the relevant one or more manufacturing variables of feature with the multiphase flow of producing by single well;

F) on according to the estimation model of step b) and basis, repeat to estimate the dynamic fluid-mixing flow pattern of well group according to the manufacturing variables of step e) monitoring;

G) carry out dynamic harmonic process, wherein, during selected mediation:

-suppose that the dynamic mixed flow flow pattern according to the estimation of step f) is the accumulative total that multiply by described each well production estimation model of unknown weight coefficient;

-unknown weight coefficient mates basically up to the dynamic fluid flow pattern of dynamic fluid-mixing flow pattern of estimating and monitoring and estimates by repeating to change each weight coefficient; And

-optimal estimation of the production stream during selected the mediation is provided, wherein use each well to be in harmonious proportion factor and estimation model and estimate the estimated output of each well during next the mediation;

H) limit the operation optimization target, this operation optimization target is made up of the target to be optimized about the output of one or more wells and/or well group.

I) output of the well effluent of adjusting well group makes and approaches optimization aim; And

J) repeating step g and i every now and then.

Alternatively, the method according to this invention can also may further comprise the steps:

-discern one or more numerical value manipulated variables at least one well in the well group; this numerical value manipulated variable can be by direct manipulation to change well yield; and, thereafter from obtaining making well manipulated variable and the flux of the well effluent of production or the relevant forecast model of variation of flow pattern and/or other features by the well surveying examination and/or at normal data and/or the estimation model that mixes the production period acquisition.The well of the manipulated variable that is not identified will have for equaling name estimates the forecast model of the steady state value of well production;

-forecast model of all wells in the well group is sued for peace so that total mixing production forecast model to be provided;

-under the guidance of single well forecast model and total mixing production forecast model, regulate the output of well effluent, so that realize described optimization aim by the well manipulated variable.

Alternatively, the method according to this invention can also may further comprise the steps:

-measurement interaction pressure, such as the pressure in the one or more production manifolds in the flow duct on the well head of the well that is connected to well group, wherein, the flux of the moving pipeline of a plurality of well streams mixes in described manifold, along with the variation of total well production, the variation of described pressure is indicated the interaction between the product stream of different wells and it is associated;

-obtain to make the variation dynamic data relevant of interaction pressure from normal mixing production and/or during produce upsetting and/or with the measurand of well by carrying out a series of interactions tests that its interphase interaction pressure changes;

-obtain to make the variation of well manipulated variable and the variation and the relevant well forecast model of well production of interaction pressure from the variation dynamic data relevant that makes interaction pressure with the measurand of well;

-from normal mixing production period and/or during produce upsetting and/or by carrying out the variation and the relevant dynamic data of total mixing output that a series of tests that its interphase interaction pressure changes obtain to make interaction pressure, and thereafter, obtain to make the variation and the relevant one or more manifold interaction models of total mixing output stream that flow through manifold of interaction pressure;

-the well forecast model is combined with the interaction pressure model so that obtain always to mix the production forecast model.

Alternatively, the method according to this invention can also comprise by revising forecast model and the step of repeated optimization method periodically according to current flowing, make forecast model reflection of revising as current the flowing of being estimated by dynamic harmonic process.

Optimization aim can be that make an accumulative total or average combination and/or single well production and the relevant revenue function of pecuniary gain of actual pure pecuniary gain or pecuniary gain or growth always, comprises relevant cost of production alternatively.

May be required in and realize optimization aim when obeying output constraints, this output constraints comprises the constraint to the output of manipulated variable and/or each well production and/or well production quantity (comprising survey data) and/or well group, and/or to the constraint of interaction pressure and/or to mixing the constraint of total output.

The method according to this invention can also comprise the step that is optimized by a plurality of numerical optimization algorithms that use with respect to manipulated variable, it is based on operation optimization target (having constraints alternatively) and well production forecast model and/or always mix the production forecast model, so that produce one group of optimum manipulated variable of realizing the operation optimization target.

Alternatively, the well head place by being adjusted in well or with flow duct that well is connected in the output choke valve open or flow control valve in the gaslift gas injection system of well open or other methods by excitation or restriction well production change the well effluent output of well.

Alternatively, can pass the well effluent output that the parallel production manifold pipeline that is connected between upstream manifold and the downstream manifold is regulated the interaction pressure of production system or changed well by the pressure of regulated fluid separation assembly (single or multiple) by well production is resend.

By the method according to this invention predict for realizing that the required adjusting of optimization aim can be sent to well and production system automatically, perhaps alternatively, after operating personnel's checking, be sent to well and production system.

Alternatively, can be partially or completely generate one or more estimations and/or forecast model according to physics theory and/or experience of well and/or production system and/or machinery and/or chemical feature.

Can regulate the variation of optimization aim with the state of the state of reflection and/or prediction production demand and/or cost and/or income and/or production basis facility and/or well and/or production equipment; And alternatively, optimization aim can be implemented by being optimized process, and the result of this optimizing process is performed and/or is used to analyze and plans and/or be recorded to be used for operation in the future.

The production of the well of producing in the first beginning and end that the method and system of this general introduction also can be applicable to the production by stopping the one or more wells in the well group alternatively temporarily or starts well group realize the situation of optimization aim.

Estimate and/or forecast model in one or morely can compare and/or estimate with respect to the physics of the theory of well and/or production system and/or experience and/or machinery and/or chemical feature alternatively; For fault diagnosis and/or analyzing and diagnosing and/or be used for improved model and/or be used for analyzing, thereby cause long period scope production management and optimization activity.

When one or more wells of well group periodically or are off and on operated or are operated every now and then, and with optimised and restrained alternatively output or relevant amount greater than periodically or the fixed time period of the time cycle of intermittently operated when being estimated (for example by average), method of the present invention also is suitable for.

When one or more wells of well group by operation periodically or off and on or when operating every now and then, with and operation duration (as the part of fixed time period) when being used as the manipulated variable of well, method of the present invention also is suitable for.

Method of the present invention is applied in addition to the optimization aim that well limited in having the well group of two or more underground payzones.In this case, except " well production estimation model " and " well production forecast model ", also generate " payzone output estimation model " and " payzone production forecast model ".

The method according to this invention allow individually and in the scope of overall manufacturing equipment with the function that can freely handle and also be used for the variable of overall device at each well place that is characterized by of well performance.The sign of well and they are predicted directly, accurately, in real time in the scope of production equipment with the interaction permission of equipment and are optimized well production.The method according to this invention can comprise the factor to the constraints of producing, and these constraintss had not only resulted from the interaction between the well that the restriction of equipment causes but also resulted from the constraints that the outside applies.The method according to this invention is also referred to as " the overall real-time optimization of output " (PU RTO).

Method according to " PU RTO " of the present invention relative prior art of method has several respects advantage, for example, and as in the described PU RTM of International Patent Application PCT/EP2005/055680, summarizing.Especially, can be used for obtaining various wells and production system feature according to " PU RTO " of the present invention method from simple well and the production test carried out separately at well and production equipment, make the easier maintenance of model and need not the survey data and the amount of discontinuous measurement, although in production environment, may change, unpredictablely such as the surface roughness of pipeline, the fluid behaviour and the variablees such as constituent, device and well performance curve of reservoir pressure-volume-temperature through a plurality of time cycles.In other words, " PU RTO " is " data-driven ".Especially, " the total well and the production system model " of combined shaft production system need not be envisioned and just can be configured aspect its basic physical property except using basic topological sum physical relation according to survey data purely.

The method according to this invention can be used to provide the well of combination and the sign of production system, and it will be of value to off-line analysis and planning activity in addition.In other words, another benefit of the present invention is that it can provide a kind of real-time current well production feature and off-line analysis and modeling are combined with the method and system of supporting with widely, the production capacity of integraterd manufacturing system is planned and optimized to mode weekly, every month, and described integraterd manufacturing system comprises a plurality of well groups and relevant production equipment.

Description of drawings

By example the present invention is described in more detail with reference to accompanying drawing, wherein:

Fig. 1 has schematically shown according to production system of the present invention, wherein produced the multiphase flow mixture that comprises crude oil, water, natural gas and/or other fluids, and this multiphase flow mixture is transported in the volume eliminator via the multiphase flow feed-line by the well group of a plurality of wells of representing with two wells;

How Fig. 2 has schematically shown from the test data of one group of well and production system and has generated " the well production forecast model of correction " and " well of correction and total output forecast model "; And

Fig. 3 how schematically to have shown will " optimization of well operation output " problem and " optimization of overall device operation output " problem be formulated so that optimize selected well and find the solution " optimum setting point " for single well, and find the solution " optimum setting point " for other wells and overall manufacturing equipment.

The specific embodiment

With reference to Fig. 1.Fig. 1 has described a simple embodiment of the production system that comprises well group, and the product of described well group is in that to produce the manifold place mixed and be sent to production separator.Shown well 1 in detail, this well 1 can be regarded the representative of other wells in the well group as.Yet other wells in the well group can be different aspect the pattern of the character of its product and flux and/or operation/excitation/manipulation.

Well 1 comprises the casing 3 in the well that is fixed on subsurface formations 4 and extends to the production pipe 5 of subsurface formations from ground.Well 1 also comprises the well head 10 that is provided with the supervising device that is used to carry out well measurements, and described well measurements is generally used for measuring tube head pressure (THP) 13 and flowline pressure (FLP) 14.Alternatively, can be useful on the pipe on the downhole monitoring device that carries out underground survey (for example down-hole pipe pressure (DHP) 18) and/or underground and/or ground and/or the differential manometer of flowline (wet gas meter (not shown) for example.Well can also be produced from a plurality of underground payzones or branch.The well head of the well in the well group can be positioned at land or sea, on the sea level or in the seabed.

In well 1 (with well 2) and the well group some but need not to be other all wells and also will have the device that some regulate output, such as: output control restriction choke 11 or fixing beam type oil nozzle restriction choke (not shown) and/or gaslift gas injection control system 12 or downhole interval control valve (not shown), the output in one or more inflows district of its control well." manipulated variable " of numerical value regulate in the device of output with these each be associated.

Production system also comprises from well head 10 and extends to a plurality of production flow duct 20 of producing manifold 21, produces pipeline 23 and separate the device (being production separator 25 in this case) that mixes multiphase flow.Produce MP meter 22 and production separator pressometer 26 as directed usually can be used on production manifold and the production separator.The device that pressure reduction between some single-phase outlets that are used to regulate the production separator level and regulate its pressure or eliminator alternatively will be arranged.For simplicity, shown pressure control loop 27 among Fig. 1.Usually, produce MP survey data 22 (alternatively, production separator pressure measurement data 26) will be as " interaction pressure ", it changes along with productivity ratio and changes, and this production MP survey data is the indicator of the interaction degree between a kind of well.

Production separator 25 is provided with and is respectively applied for water, the gentle outlet 35,36 and 37 of oil.In the outlet 35,36 and 37 each is respectively equipped with flow measurement device 45,46 and 47.Alternatively, water and oil outlet can be combined in together.Can be alternatively by regulating pressure from 37 throughput control production separator, thereby influence MP 26 and flowline pressure 14 and so influences each well yield.

The well measurements data comprise at least the data from 13, optionally from 14,18 data, the gaslift charge velocity from 12, the position of output restriction choke 11 and the data of other available devices, these well measurements data are continuously delivered to " production data acquisition and control system " 50.Similarly, mix output survey data 45,46,47 and also be continuously delivered to " production data acquisition and control system " 50.Common data transfer path is represented with 14a and 45a.Data in 50 are stored, and can be used for the non-real-time data retrieval subsequently, are used for data analysis and structure model as summarizing at this patent.Data in " production data acquisition and control system " are used in combination with " well production estimation model " estimated in real time continuously that be used for single well production also by " PU RTM " real time access.Also can regulate some well productivity ratio controls (for example, produce restriction choke and open or the gaslift charge velocity) according to " production data acquisition and control system " 50, be used for remote adjustment and optimize well production, the holding wire of gaslift charge velocity control is shown as 12a.

Relevant well test facility can be alternatively and is preferably used for the sign of single well surveying examination and a plurality of wells.When not having well test facility, can be used to carry out the structure of well surveying die trial type from the survey data 45,46,47 of producing eliminator.

Referring now to Fig. 2, Fig. 2 provides a preferred embodiment of the modeling process of " data-driven " of the present invention.Purpose is only to have considered the relevant especially feature of well and production system and effect and generate the enough useful model that is fit to purpose of the present invention.This program has generated and has been used for n well i=1, and 2 ..., " the well production forecast model of correction " of the well group of n and " well of correction and total output forecast model ", wherein, " manipulated variable " of each well is identified, and special-purpose well equipment is arranged, and is described below:

-carry out a series of well surveying examinations, wherein each well yield changes during this period, is used for coming well is characterized by " manipulated variable " that change this well.Well surveying examination data 60 are used for y _i=f _i(u _i, v _i) form generate " well production estimation model " 61, for u _i, v _iBe in group U _i* V _iEffective in the scope, wherein vectorial y _iBe the gentle output of oil, water of well i, u _iBe measurement vector at JingiChu, v _iIt is manipulated variable at JingiChu.Use process that special-purpose well test facility makes up " well production estimation model " with before in " PU RTM ", summarized the same.Measurand u in " PU RTM " _iWith manipulated variable v _iDo not distinguish between (it is also measured), but need in the present invention to distinguish;

-carry out a series of " well and production equipment interact and test " then so that obtain " well and production equipment interaction test data " 62, perhaps alternatively, data also can obtain from " creation data " 63, and described " creation data " is included in the dynamic event record during the ordinary production (such as when main gas production well is closed).

-structure the variation that makes the variation of " manipulated variable " of well and " interaction pressure " and relevant " well production manipulation and the interaction model " 65 of survey data then at the flow duct place of well.Relation between the survey data at the moving pipeline place of " manipulated variable " and well stream can by regulate manipulated variable at the normal well production period and be recorded in the moving pipeline place of well stream survey data reaction and obtain from well surveying examination data (for example " DDWTs ") 60 or from " creation data " 63.Similarly, the relation between the survey data at the moving pipeline place of " interaction pressure " and well stream can be by setting up according to " creation data " 63 or " the interaction test of well and production equipment " data 62 of being recorded in the dynamic event during the ordinary production.

For each well i=1,2 ..., " well production is handled and interaction model " 65 forms that will have of n are u _i=g _i(v _i, w), wherein w ∈ W is shared " the interaction pressure " in group W scope, u _iBe measurement vector at JingiChu, v _iBe manipulated variable at JingiChu, the same with before.

Then, each well i=1,2 ..., " the well production forecast model " 66 of n will generate according to " well production estimation model " 61 and " well production is handled and interaction model " 65, be expressed as follows:

y _i＝f _i(u _i，v _i)＝f _i(g _i(v _i，w)，v _i)＝h _i(v _i，w)，

So that function h _i(v _i, w) make vectorial y _i(oil of well i, the gentle output of water) and w (shared " interaction pressure ") and v _i(manipulated variable of JingiChu) is relevant.

The sign of given well 66, need now to obtain " interaction pressure " w at the production manifold place of present embodiment with always mix the correlation of producing flow via what produce the manifold transmission, and alternatively at the variable v of the overall device level pressure set-point 27 of production separator (for example, the) lower-pilot of also influence " interaction pressure " _wCorrelation.This sets up " interaction pressure model " 67 by using " well and production equipment interact and test " data 62, $w=k(\hat{y},v_w),$ And realize, wherein,

Be defined by sending to the vector of the summation of the well production of producing manifold, in this case,

$\hat{y}=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{y}_i.$

There is following situation, wherein unavailable or unreliable for the certain well or the part well of whole system from the data of well surveying examination or output, for example when well does not put into production as yet.In this case, the model of representing based on physics theory and/or experience of well and/or production system and/or machinery and/or chemical feature can be used for replacing 65 or 67.

In case generate " well production is handled and interaction model " 66 and " interaction pressure model " 67, according to as the online part of the flow process of Fig. 2 and then as the flow performing optimizing process of Fig. 3.The output of wishing is optimized expression formula and is suggested as follows:

The well optimization 78 usefulness subscript group I of-selected well (a suitable subclass 1.2 ..., the form that n) is defined as is

$\underset{v_i}{\max }{R}_i(y_i,v_i)$

The constraints that is stood is c _{I, j}(y _i, v _i, u _i) 〉=0, j=1,2 ..., J _iWherein, R _i(y _i, v _i) be target or the revenue function 76 of each well i, by changing v _i(manipulated variable of JingiChu) and be maximized, and stand respectively about y _i(well production), v _i(well manipulated variable) and u _iUsing of (measurand) retrains inequality c _{I, j}(y _i, v _i, u _iThe J that) 〉=0 limits _iConstraints 77.

-for well $i\∉I$ The form of (not as the optimised well of a part of 78) total output optimization 83 is

$\underset{v_i,i\∉I,v_w}{\max }{R}_i(\hat{y},v_i,u_i,w,v_w)$

The constraints that is stood is $c_j(\hat{y},y_i,v_i,u_i,w,v_w)\≥0,$ j＝1.2，...，J _i。Wherein,

Be target or revenue function 81, by changing v _i(well $i\∉I$ The manipulated variable at place) and be maximized, and stand respectively about (always mixing output), w (interaction pressure) and v _w(the optionally vector of the variable that under total equipment level, can handle) and y _i(single well production), v _i(well manipulated variable) and u _i(measurand) i=1.2,3 ..., the J of n _wConstraints 82, $c_j(\hat{y},y_i,v_i,u_i,w,v_w)\≥0.$

For being in the well that normal mixing is produced, " PU RTM " on-line operation is estimated so that produce real-time continuously output at each single Jing70Chu.Single well production

" current " " PURTM " estimate and " current " survey data and manipulated variable value

Be used for by being limited to

With

The constant offset at place $d_i={\tilde{y}}_i-h_i({\tilde{v}}_i,\tilde{w})$ Revise " well production is handled and interaction model " y _i=h _i(v _i, w)." the well production forecast model of correction " 75 forms that have are y then _i=h _i(v _i, w)+d _iMakeover process allows optimizing process to concentrate on the incremental variations of well production and total output, if make ${\mathrm{\Δv}}_i=v_i-{\tilde{v}}_i,$ $\mathrm{\Δw}=w-\tilde{w},$ ${\mathrm{\Δy}}_i=y_i-{\tilde{y}}_i,$ Then if Δ v _i=0 and Δ w=0, Δ y so _i=0 _iIllustration purpose presented for purpose of illustration can be with the form Δ y of " dividing deviation (separated difference) " for " the well production forecast model of correction " 75 of each well i _i=A _iΔ v _i+ B _iΔ w _iCalculate.Symbol A _iB _iCan be regarded as or matrix or operation Δ v _iFunction with Δ w.Alternatively, Δ v _iWith the cross-term of Δ w and second and more high-order term can be inserted into and not lose generality.

Carry out each well i ∈ I in the well that well optimizes for hope, given its " well production forecast model of correction " 75 and target or revenue function 76 and relevant optimization constraints 77 then can make up well optimization 78 so that find the solution optimal value v _i, 79 (manipulated variables of JingiChu).It should be noted that well optimization must suppose that shared interaction variable w is constant or can ignore the influence of optimizing the result when well is optimized, this variable w is one and is subjected to the common output of well and the variable that the variable under total production system level influences.

" interaction pressure model " 67, $w=k(\hat{y},v_w),$ Can be by limiting steady state value $d_w=\tilde{w}-k(\tilde{y},{\tilde{v}}_w)$ And be corrected, wherein, mix output and measure

" current nominal value " use

Expression.Then, the form of " the interaction pressure model of correction " is $w=k(\hat{y},v_w)+d_w.$ In addition, we suppose to estimate

Mix the output test and fully be in harmonious proportion, make by " PU RTM " and nearest name are total

$\tilde{y}=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\tilde{y}}_i$

Moreover, be in the example illustration purpose, " the interaction pressure model of correction " also can have the form of " branch deviation " $\mathrm{\Δw}=\mathrm{K\Δ}\hat{y}+\mathrm{L\Δ}{v}_w,$ Wherein

$\mathrm{\Δ}\hat{y}=\hat{y}-\tilde{y}=\hat{y}-\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\tilde{y}}_i$

And $\mathrm{\Δ}{v}_w=v_w-{\tilde{v}}_w.$

Now

It is estimation from total mixing output of " revising the well production forecast model " 75.Therefore, " well of correction+total output forecast model " 80 of " difference " form is by with " the well production forecast model of correction " 75 and " the interaction pressure model of correction " combined being configured.From

$\mathrm{\Δ}\hat{y}=\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{y}_i\right)-\tilde{y}=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\Δy}}_i=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{A}_i{\mathrm{\Δv}}_i+B_i\mathrm{\Δw}=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{A}_i{\mathrm{\Δv}}_i+\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{B}_i\right)(\mathrm{K\Δ}\hat{y}+\mathrm{L\Δ}{v}_w)$

We obtain

$\mathrm{\Δ}\hat{y}=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{A}_i{\mathrm{\Δv}}_i+\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{B}_i\right)(\mathrm{K\Δ}\hat{y}+\mathrm{L\Δ}{v}_w)$

It is respectively with variable

(always mixing output), Δ v _i(manipulated variable of Jing Chu), Δ v _w" well of correction+total output forecast model " 80 of the implicit function form that (manipulated variable under total production system level) is relevant.For Δ v _i, Δ v _wSet-point, according to functional form

$\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{A}_i,\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{B}_i,K,L,$

80 implicit function form

Can find the solution by several different methods.In this case, interaction composition

$\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{B}_i,K,L$

Be real number matrix, for example, we have

$[I-\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{B}_i\right)K]\mathrm{\Δ}\hat{y}=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{A}_i\mathrm{\Δ}{v}_i+\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{B}_i\right)\mathrm{L\Δ}{v}_w$

For given v _i, v _wValue, it can be found the solution

Unless, operator for example

$[I-\left(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{B}_i\right)K]$

Be irreversible.

Then, optimize the output necessary operations variable v of overall device _i, v _wEstimation by " well of correction+total output forecast model " 80 and target or revenue function 81 and relevant optimization constraints 82 combinations are obtained, so that form " optimization of total equipment operation output " 83.Then, be optimized so that find the solution " optimum performance variable " (v _i, v _w) optimal value 84.

Form according to 83 is optimized variable 84 and can directly be calculated or use automatic Iteration Optimization program.The multiple automatic Iteration Optimization method that can use according to 83 form is arranged.For example, M Bazaraa referring to publication in 1993, the textbook that H.D.Sherall and C.M.Shetty write " non-linear programming-theory and algorithm (NolinearProgramming-Theory and Algorithms) " second edition, the strictness of other types or didactic " global optimization " method, for example be published in " computer and Chemical Engineering " referring to L.T.Biegler and I.E.Grossmann, 2004, the 28th phase, " first: the review of optimization " of 1169-1218 page or leaf and " second portion: the prospect of optimization " are hereby incorporated by.For manipulated variable wherein is the preferred embodiment that continuous variable and 83 limits by smooth continuously nonlinear model and revenue function and inequality constraints condition, and the continuous quadratic equation program (SQP) that has a plurality of starting points is used to carry out automatic Iteration Optimization so that obtain " optimum manipulated variable ".

Then, " optimization manipulated variable " group can be used for further operation.Alternatively, " optimization manipulated variable " is reported to the production equipment operator so that in well and equipment place executable operations, perhaps alternatively, directly is sent to " production data acquisition and the control system " 50 that is used to automatically perform operation.

Be optimized the calculating and the application of manipulated variable every now and then, and it is by " optimization start-up system " 90 controls.Preferably, " well operation output optimization " and " total equipment is operated output optimization " periodically (for example, once a day) be activated and/or anticipate well management philosophy or production system or constraints or optimization aim state change and be activated according to demand.In one embodiment, gaslift can will start optimizing process automatically with the variation of situation.

In the preferred embodiment of " PU RTO " according to the present invention method:

-use current latest data and historical test data to verify where necessary and upgrade all models.

-periodically verify by checking the normal well test and upgrade " PU RTM well production estimation model ".

-well manipulated variable is periodically circulated so that allow checking and upgrade " well is handled and interaction model " at the normal well production period.

-under normal operating condition, obtain yield data so that verify when needed and upgrade " well operation and interaction model " and " interaction pressure model ".

-check back transmission " optimization manipulated variable " the person of being operated.

-calculation optimization step is also determined or is opened so that recover single well production or be closed so that stop single well production, the perhaps well of changing between different production separators except " the optimization manipulated variable " of the well determining to produce.In this case, manipulated variable v _i, v _wTo comprise binary numeral 0 or 1, and can and the optimizing expression formula according to used model structure with didactic method and be used for finding the solution of different strictnesses.

The well of the actual conditions of-reflection well and production system and the forecast model of total production system should be periodically compare and estimate with the physics and the mechanical model (if any) of the theory of well and/or production system.The model that is obtained from real well performance according to the present invention will produce the information of help long period scope production management and optimization activity with respect to the evaluation and the contrast of theoretical model.

Claims (18)

1. method that is used to optimize well group output, the well effluent stream of described well group mixed and in fluid separating assembly, be separated into crude oil, gas and/or other fluids to the small part separated flow, described method comprises:

A) each well is carried out well surveying examination, during this period, tested well yield be changed and the manufacturing variables of one or more single wells monitored;

B) according to the data that obtain by well surveying examination obtain with by by the estimation model of each relevant well of the variation of the well manufacturing variables of the variation of the flow pattern of the product of well logging production and monitoring;

C) make normal miscella of well group input and/or gas production;

D) during step c), the dynamic fluid flow pattern to the small part separated flow of flow meter monitoring crude oil, gas and/or other fluids to the small part separated flow of crude oil, gas and/or other fluids by being arranged in the fluid separating assembly downstream;

E) during step c), monitor the relevant one or more well manufacturing variables of feature with the multiphase flow of producing by single well;

F) on according to the estimation model of step b) and basis, repeat to estimate the dynamic fluid-mixing flow pattern of well group according to the manufacturing variables of step e) monitoring;

G) carry out dynamic harmonic process, wherein, during selected mediation:

-suppose that the dynamic mixed flow flow pattern according to the estimation of step f) is the accumulative total that multiply by described each well production estimation model of unknown weight coefficient;

-unknown weight coefficient mates basically up to the dynamic fluid flow pattern of dynamic fluid-mixing flow pattern of estimating and monitoring and estimates by repeating to change each weight coefficient; And

-optimal estimation of the production stream during selected the mediation is provided, wherein use single well to be in harmonious proportion the factor and estimation model and estimate the estimated output of each well during next the mediation;

H) limit the operation optimization target, described operation optimization target is made up of the target to be optimized about the output of one or more wells and/or well group;

I) output of the well effluent of adjusting well group makes and approaches optimization aim; And

J) repeating step g and i every now and then.

2. method according to claim 1, further comprising the steps of:

-discern one or more numerical value at least one well in the well group to handle the well manufacturing variables, described numerical value is handled the well manufacturing variables can be by direct manipulation to change well yield, and, thereafter from obtaining making well manipulated variable and the flux of the well effluent of producing or the relevant forecast model of variation of flow pattern and/or other features by the well surveying examination and/or at normal data and/or the estimation model that mixes the production period acquisition, wherein, not being identified any well of handling the well manufacturing variables will have for equaling the nominal forecast model of estimating the steady state value of well production;

-forecast model of all wells in the well group is sued for peace so that total mixing production forecast model to be provided;

-under the guidance of single well forecast model and total mixing production forecast model, regulate the output of well effluent, so that realize described optimization aim by the well manipulated variable.

3. method according to claim 2, further comprising the steps of:

The one or more interaction pressure of-measurement, such as the pressure in the moving ducted one or more production manifolds of the well stream that is connected with the well head of the well of well group, wherein, the flux of the moving pipeline of a plurality of well streams mixes in described manifold, along with the variation of total well production, the variation of described pressure is indicated the interaction between the product stream of different wells and it is associated;

-obtain to make the variation dynamic data relevant of interaction pressure from normal mixing production and/or during produce upsetting and/or with the test variable of well by carrying out a series of interactions tests that its interphase interaction pressure changes;

-obtain that the well manipulated variable is changed and the variation of interaction pressure and the relevant well forecast model of well production from the variation dynamic data relevant that makes interaction pressure with the measurand of well;

-obtain to make the variation of one or more interaction pressure to mix the relevant dynamic data of output from normal mixing production period and/or during produce upsetting and/or with total by carrying out a series of tests that its interphase interaction pressure changes, and total mixing output that acquisition thereafter makes the variation of one or more interaction pressure and flows through manifold flows relevant one or more manifold interaction models;

-the well forecast model is combined with the interaction pressure model so that obtain always to mix the production forecast model.

4. according to claim 2 and 3 described methods, described method also comprises repeated optimization method periodically, and it is by revising current the flowing that forecast model makes that the forecast model reflection of correction is estimated by dynamic harmonic process according to current flowing.

5. according to each described method among the claim 1-4, wherein, optimization aim is that make an accumulative total or average combination and/or single well production and the relevant revenue function of pecuniary gain of actual pure pecuniary gain or pecuniary gain or growth always, comprises relevant cost of production alternatively.

6. method according to claim 5, wherein, when obeying output constraints, realize optimization aim, described output constraints comprises manipulated variable and/or each well production and/or comprises the well production quantity of survey data and/or the constraint of the output of well group, and/or to the constraint of interaction pressure and/or to the constraint of the mixing total output of well group.

7. according to claim 5 or 6 described methods, also comprise by using any step that is optimized in a plurality of numerical optimization algorithms of manipulated variable, it is based on operation optimization target that has constraints alternatively and well production forecast model and/or always mix the production forecast model, so that produce one group of optimum manipulated variable of realizing the operation optimization target.

8. according to each the described method among the claim 1-7, wherein, the well effluent output of well changes in the following manner, promptly the well head place by being adjusted in well or with flow duct that well is connected in opening of output choke valve or opening or device by other excitations or restriction well production of flow control valve in the gaslift gas injection system of well, such as any reversible and the controlled device of closing and open well, have set point at Jing Chu as the control loop of the output choke valve of actuator, lift charge velocity or pressure about gas well gas, gas well gas is lifted to inject and is closed the duration, gas well gas is lifted to inject and is opened the duration, gas well gas is lifted the set point of the control loop of charge velocity, the control loop on well water jet pump hydraulic fluid injection circuit, the speed of well electric submersible water pump (ESP), well insert pump motor speed, the set point of opening of the downhole interval control valve of closing duration and/or well of well insert pump.

9. according to each the described method among the claim 1-8, wherein, the well effluent output of well changes in the following manner, promptly passes the parallel production manifold pipeline that is connected between upstream manifold and the downstream manifold and regulates one or more interaction pressure of production system by utilizing well production resend, and/or the pressure by the regulated fluid separation assembly, and/or be used for well effluent sent to by adjusting and mix output and maybe will mix the valve that output sends to one or more manifolds of one or more production separators, and/or the speed by the compressor in the output channel of regulated fluid separation assembly.

10. according to each the described method among the claim 1-9, wherein, be sent to well and production system automatically, alternatively, after operating personnel's checking, be sent to well and production system for realizing the required adjusting that optimization aim is predicted.

11. according to each the described method among the claim 1-10, wherein, one or more estimations and/or forecast model can be alternatively generate according to physics theory and/or experience of well and/or production system and/or machinery and/or chemical feature partially or completely.

12. according to each the described method among the claim 1-11, wherein, optimization aim is conditioned the variation with the state of the state of reflection and/or prediction production demand and/or cost and/or income and/or production basis facility and/or well and/or production equipment; And alternatively, optimization aim can be implemented by being optimized process, and the result of described optimizing process is performed and/or is used to analyze and plans and/or be recorded to be used for operation in the future.

13. according to each the described method among the claim 1-12, wherein, optimization aim by stopping the one or more wells in the well group alternatively temporarily production or the production that starts the well that first beginning and end of well group produce realize.

14. according to claim 2 and 3 described methods, wherein, estimate and/or forecast model in one or morely can compare and/or estimate with respect to physics theory and/or experience of well and/or production system and/or machinery and/or chemical feature alternatively; For fault diagnosis and/or analyzing and diagnosing and/or be used for improved model and/or be used for analyzing, cause long period scope production management and optimization activity.

15. according to each the described method among the claim 1-13, wherein, one or more wells in the well group are by periodically or off and on operation or operation every now and then, and will optimised with restrained alternatively output or relevant measuring at fixed time period greater than time cycle of periodicity or intermittently operated estimated (for example by on average).

16. method according to claim 15, wherein, one or more wells of well group are by periodically or off and on operation or operation every now and then, and are used as the manipulated variable of this well as its operation duration of the part of fixed time period.

17. according to each the described method among the claim 1-13, wherein, described method is applied in addition to the optimization aim that well limited in the well group that has two or more underground inflow districts, under described situation, except " well production estimation model " and " well production forecast model ", also generate " payzone output estimation model " and " payzone production forecast model ".

18. method according to claim 1, well manufacturing variables wherein comprises one or more following variablees such as pressure and/or other fluid flow characteristics of each well effluent stream: well casing head pressure; Well stream moves pipeline pressure; Well casing head temperature; Well stream moves pipe temperature; Cross over well and produce the pressure reduction of choke valve; Leap on the moving pipeline of well stream comprises the pressure reduction of arbitrary pressure reduction generator of wet type gas Venturi tube; The flow measurement data, nominally such as the flow measurement data that are only applicable to single-phase flow, described flow measurement data are suitable for use as the input of well estimation model, even also are like this when Dang Jing has multiphase flow; Well is produced the open mode or the position of choke valve; The open mode and the position of reversible and controlled any device of closing and opening well; Gas well gas is lifted charge velocity; The hydraulic fluid charge velocity of well water jet pump; Well production casing pressure; The speed of well electric submersible water pump ESP; The inlet pressure of well ESP pump; The blowdown presssure of the down-hole pump of well ESP; The pressure reduction of the down-hole Venturi tube of well ESP; The power consumption of well ESP; The motor phase current of well ESP; The motor input power of well insert pump; The motor speed of well insert pump; Well rod-type throw of pump discharge capacity; The load unit data of well insert pump; The position of beam pump gearbox shaft; Speed difference/the motor of well insert pump/gear-box slides; The down-hole pipe pressure of well; The down hole annular space pressure of well; The down-hole pipe of well or the temperature of well annular space, or it is from the different derivative datas of distributed temperature sensor; The open mode of the downhole interval control valve of well; Amplitude from the sound frequency selection that is installed in the one or more sonic transducers on the moving pipeline of well stream; Swim over to the propagation delay situation of the relevant sound pattern that the frequency of two or more sonic transducers of installing on the downstream direction selects on the moving pipeline of well stream.

Images