CN101361080B - Method for oil gas field large-scale production optimization - Google Patents

Abstract

A method for enhancing the allocation of fluid flow rates among a plurality of well bores in fluid communication with at least one subterranean reservoir is disclosed (24). An objective function and system equations are generated which utilize constraint violation penalties associated with soft constraints. The soft constraints are constraints which may be violated if necessary to arrive at a feasible solution to optimizing the objective function and the system equations. The fluid flow rates are then allocated among the well bores (30) as determined by the optimizing of the objective function and system equations. Fluid flow rates among well bores (30), particularly those exhibiting similar fluid characteristics, may be related to one another. Initial flow rates of components (oil, gas, and water) and pressures in the well bores (30) may be determined by an initial simulation run.

Description

The method of oil gas field large-scale production optimization

Technical field

The present invention relates generally to be used to control the method for the production of hydrocarbons of oil gas field, and relate more particularly to be used for come the method for optimization production by increasing flow distribution of fluid among the well.

Background technology

The oil gas field scale optimization of the output of the known production fluid (comprising hydrocarbon) of attempting to optimize or increase the oil gas field that comes self-contained one or more subsurface reservoirs.Well or well connect reservoir with uphole equipment, and the production fluid of being caught is collected and handled to this uphole equipment.Typically, these produce the component that fluid comprises oil, gas, water.Flow controller or volume control device are used for regulating the assignment of traffic among the well in well site.Can by regulate flow controller control the oil of single well, gas, water different component relative turnout and produce than so that change pressure in the well.

Need uphole equipment to produce and the process for producing fluid.These equipment can comprise the device such as separation vessel, pump, storage tank, compressor etc.Ideally, by using minimum as far as possible and the most cheap uphole equipment to minimize the capital outlay of these equipment.Yet the fluid treatment ability should be enough big so that excessively do not limit the oil of economic needs and/or the production flow of gas.Therefore, the distribution of fluid stream is optimized to the recovery of maximization money ideally in the well, satisfies the production constraint that is applied such as the fluid treatment ability by uphole equipment simultaneously.

Optimisation technique is used for predicting with one group of given production constraint the optimal allocation of the fluid stream of well.At first, the reservoir simulation device is used for the fluid stream that mathematical simulation runs through the oil gas field that comprises reservoir and well.Analog stream is used to set up the composition flow rate curve or the flow equation of each well, and how relevant its flow of describing a kind of component (such as water) is with the flow of another kind of component (for example oil).Typically, set up and to manage optimization such as maximum oil production or minimize the objective function of the target of aquatic product.Objective function merges the flow from the well of being predicted by reservoir simulation.Specify one group to produce constraint, such as the restriction of production of oily productive target or the gas or the water of oil gas field.Form equation of constraint and produce constraint to satisfy these.Fluid stream among the well must depend on these and produce constraint.Then objective function by the subroutine optimization that is called as the optimizer device so that determine the optimal allocation of the flow among the well.The optimizer device utilizes well composition flow rate equation and equation of constraint in optimizing process.

First shortcoming of typical oil gas field scale optimization design is the production constraint that the feasible solution of optimization can not be used to specify.For example, the oil yield of certain level may be needed, but water can not be produced more than specified amount.The feasible solution that has the objective function of this group constraint is impossible.Under this occasion, must adjust one or more constraints, and move reservoir simulation device and optimizer device once more so that when determining that feasible solution is possible.The iteration of moving in a large amount of optimizations of processing target function is calculating intensive (intensive) and undesirable.

Though second problem in some optimal design is the feasible solution that can obtain objective function optimization, yet its possibility of result is unpractical.For example, in for the first time operation or time step, the optimizer device can determine that first well should high level produces and second well is closed basically.In next time step, the optimizer device can point out second well high level to produce and first well is closed basically.So if follow from the suggested distribution of optimizer device, the production of well may be vibrated.Usually, if be in consistent level from the production of the well with similar fluid mobility, then it is more practical.During time step, this will minimize the vibration in the production of auto-correlation well.

The 3rd shortcoming be generate the composition flow rate curve of wellbore fluid production or equation can be calculate intensive.A kind of method of calculating these flow curves be when flow controller be opened and reservoir and well between pressure drop (pressure draw down) when increasing, set up the well and the production flow of the submodel of reservoir and iterative processing component (for example oily, gas and water) on every side.Typically, must carry out several times Newton iteration producing each data point, described data point set up in well, give constant pressure drop next plant the Relationship with Yield of component with respect to another kind of component.In addition, the pressure drop in the well is relevant with how opening of the flow controller of controlling well.Repeatedly repeat this process up to enough as calculated data points, can be, make to form total-flow-rate curve or equation as 30-50 data point.Then, optimizer device use traffic curve or equation during objective function optimization.Use these repeatedly Newton iteration set up flow curve or equation and the data point calculation expense height that produces.

The invention provides solution at the above-mentioned shortcoming of traditional oil gas field scale optimization design.At first, generate objective function and related equation of constraint, it can be processed to produce feasible solution in the single run of optimizer device.Secondly, can generate equation of constraint, described equation of constraint need be from the production flow of similar well to be associated so that prevent the remarkable vibration of the well yield between the time step of reservoir simulation.At last, describe to produce the well composition flow rate curve that the production flow between the fluid components with well is associated or the effective ways of equation.

Summary of the invention

The present invention includes a kind of be used to increase with a plurality of wells that at least one subsurface reservoir fluid is communicated with among the method for flow distribution of fluid.Use numerical value reservoir simulation device at least one subsurface reservoir and with many wells that the subsurface reservoir fluid is communicated with in the simulation fluid flow.Form the composition flow rate equation by the analog stream in the well.Select tape has the production constraint of at least one the production constraint that is soft-constraint ideally, if described soft-constraint must provide feasible solution during optimizing process, then can violate this soft-constraint.Also formed corresponding to the equation of constraint of producing constraint.

Formation is corresponding to the objective function of the stream of the fluid in the well.Objective function also can comprise the constraint violation punishment corresponding to soft-constraint and soft-constraint equation.Then, utilize composition flow rate equation and equation of constraint to come the optimization aim function so that the distribution of the increase of the fluid flow among definite well.If desired, soft-constraint can be violated so that obtain the feasible solution of objective function optimization.The existence of constraint violation punishment allows soft-constraint to be violated, and still satisfies corresponding equation of constraint simultaneously.Then, among well, distribute the fluid flow of determining as by the optimization of objective function.

If desired, can preferentially to be become in the soft-constraint should be that the constraint of the most difficult violation is to obtain the feasible solution of objective function optimization to soft-constraint.The weighting scale factor can be related with the constraint violation punishment in the objective function.

The weighting scale factor can be weighted according to the priorization of soft-constraint, makes the soft-constraint of higher priority more be difficult to be violated than the soft-constraint of lower priority.

Flow between the well of selecting can have its relevant flow, especially, shows that the well of similar flowability (such as output gas oil ratio (GOR) or water-oil factor (WOR)) can have the well yield that is relative to each other.In addition, can form the equation of constraint of these relevant well flows.Then, the increase of the flow among the relevant well distributes and will be relative to each other or be bonded to each other.

In another aspect of the present invention, the well that simulated comprises a plurality of well completion unit and reservoir or comprises the reservoir of a plurality of reservoir units.Fluid stream because the pressure drop between reservoir units and well completion unit, reservoir simulation device are moved in definite reservoir units and in the well completion unit of the pressure in the well completion unit and definite at least two components (being oil and water).Then, in the scope of the fluid of each well stream, form the data point of fluid stream ratio of component.Ideally, based on determined by primary simulation device operation and with reservoir and well completion unit between the relevant composition flow rate of increase scope of pressure drop, form data point by the fluid stream in the well completion unit that converts in proportion and sue for peace.

An object of the present invention is to provide a kind of method, wherein form the objective function that comprises corresponding at least one constraint violation punishment of soft-constraint, described soft-constraint allows objective function optimised, wherein, if desired, soft-constraint can be violated so that draw the feasible solution of optimization.

Another purpose is to form the objective function that the constraint violation punishment with weighting combines, described constraint violation punishment can be weighted appropriately so that order that can priorization is violated soft-constraint.

Another purpose be in optimization, make the production flow of well relevant make finish optimize after flow among those wells have relevant flow, cause the limited oscillation of flows of those wells between the time step in reservoir simulation.

In addition, another purpose is to form the composition flow rate equation.Described composition flow rate equation is based on the variation range of initial pressure traverse figure in reservoir simulation determined flow in service and well, forms by the composition flow rate in the single well completion unit that converts in proportion.

Description of drawings

According to following explanation and unsettled claims and accompanying drawing, will understand these and other purposes, features and advantages of the present invention better, wherein:

Fig. 1 is the synoptic diagram that comprises the exemplary production of hydrocarbons oil gas field of subsurface reservoir, and described subsurface reservoir is connected to oil gas field ground by wellbore fluid, and flow controller is used to control the pressure and the flow of well, makes that the production of oil gas field can be optimised;

The process flow diagram of the illustrative methods of the oil gas field scale optimization that Fig. 2 is according to the present invention to be carried out;

Fig. 3 A and 3B have illustrated, and the intensive formed composition flow rate curve of Newton iteration method is calculated in " flow fast " formed composition flow rate curve of method used according to the invention and use;

Fig. 4 A and 4B show that well yield has the line chart of how to be correlated with between a pair of well of similar flowability;

Fig. 5 A-D illustrated when closing the pressure curve that allows in the well owing to the simulation of well flow controller when increasing, from the oil of the single well completion unit of well, the production of G﹠W;

Fig. 6 illustrated during being used for calculating the composition flow rate curve that produces well, the pressure curve of adjacent reservoir be maintained fixed constant in, allow borehole pressure curve negotiating pressure to change " c " and change;

Fig. 7 A and 7B have illustrated and have been used for the corresponding line segment and a pair of line segment of structural segmentation linear function;

Fig. 8 has shown the process flow diagram of the method for the preferred amount breakpoint that is used to select to form piecewise linear function;

Fig. 9 has described and has demonstrated breakpoint and should fall into first quartile so that prevent the chart of negative ratio;

Figure 10 has illustrated piecewise linear function.

Embodiment

Completion provide reservoir 22 and 24 and well 30,32 and 34 between fluid be communicated with.Well 34 only is connected with last reservoir 22.

Flow controller or well control system device 54,56 and 60 are used to control the fluid inflow and flow out corresponding well 30,32 and 34.Describe more fully as following, flow controller 54,56 and 60 is also controlled the pressure traverse of each well 30,32 and 34.Although do not show that well 30,32 will be connected with uphole equipment (such as oil/gas/separator, compressor, storage tank, pump, pipeline etc.) fluid with 34.The flow of the fluid stream by well 30,32 and 34 can be by the fluid treatment capabilities limits of these uphole equipments.

Fig. 2 has shown the process flow diagram that is used to illustrate according to the employed general step of oil gas field scale optimization method of the present invention.Based on the instruction that is included in the instructions of the present invention, the technician in reservoir simulation field can easily develop the computer software that is used for the method that execution graph 2 outline.

The reservoir simulation device is used for simulating the fluid flow (step 110) of the oil gas field 50 that comprises reservoir and well.Usually, this reservoir model comprise thousands of or even millions of discrete units so that carry out numerical simulation.These discrete units comprise reservoir units and well unit.The well unit comprises be used for transmitting back and forth the special well completion unit of fluid and other well unit that is communicated with flow controller and uphole equipment (not shown) fluid between adjacent reservoir units.

Specify initial and boundary condition in the oil gas field model, these initial and boundary conditions comprise, in the mode of example but be not limited to original pressure in reservoir units and well unit and flow, fluid components, viscosity etc.

Next, in the oil gas field model, carry out reservoir and the fluid flow characteristics of dry run (step 120) with step-length computing time.Especially, reservoir that is determined and the fluid flow between the well are the pressure in reservoir and the well unit.Production wellbores will receive production fluid from reservoir (comprise oil, water gentle), and described production fluid is transported to the uphole equipment of oil gas field.Inject well can be used for pressurizeing one or more reservoirs and/or be used for water treatment.Gas also can be injected in the well so that provide gas to assist (gas assisted) fluid production.Those skilled in the art will recognize available reservoir simulation device and simulate other operation that many influences are produced, and these operations within the scope of the present invention.

The component flow flow can be determined according to the stream of oil, G﹠W.Alternatively, the fluid components of stream to be optimized can be to form component, such as light hydrocarbon (C ₃-C ₄), medium hydrocarbon (C ₅-C ₈) and heavy hydrocarbon (＞C ₉).Mode by example but be not limited to, other possible combination of components can comprise that nonhydrocarbon is (such as H ₂S and CO ₂).

Next calculate the composition flow rate equation (step 130) of each well.The fluid component that these composition flow rate equations have been described in the desired extent of well flow flows with respect to the estimation of another fluid components.In the practice, the flow controller on the well can be opened or closed so that increase or reduce and export or input with respect to total fluid of well.Because the pressure traverse of the variation in the well, oil, the comparing of G﹠W of producing from well can change along with opening or closing of flow controller

The example that has shown the composition flow rate curve of well among Fig. 3 A and the 3B.In Fig. 3 A, having drawn with MSCF/D (million cubic feet/sky) is that the gas of unit is produced flow with respect to the oil production flow that is unit with STB/D (storage tank/sky).In Fig. 3 B, having drawn is that the aquatic runoff yield of unit is produced flow with respect to the oil that is unit with STB/D with STB/D.The airshed production flow of oil relatively is linear relatively in the wide region of possible oil production flow.Yet aquatic runoff yield oil production flow relatively is non-linear.The water that is produced with the high oil turnout is far more than with low oil production water that flow was produced.High throttle positions of producing flow corresponding to wide opening.

In a preferred embodiment of the invention, use " flow fast " method to produce single composition flow rate data point, it is used for constructing fast line chart or forms the composition flow rate equation then.Will be described below more detailed " flow fast " method.Person of skill in the art will appreciate that and can use additive method to produce assessment: how the component production of another component relatively may change in the total production scope of well.

The user will specify employed production constraint (step 140) together with the oil gas field model.Mode by example but be not limited to, the example of producing constraint comprises: (1) produces oil with target level; (2) produce gas with target level; (3) with the gas restriction of production under predetermined threshold; (4) aquatic product is limited under the predetermined threshold; (5) water is injected the amount of closing that is restricted to the water of producing from well; And (6) are limited on the predetermined threshold gas injection to provide gas to assist lifting.In addition, these targets and restriction can be made up mutually or be proportional.

Produce constraint and can comprise hard or soft-constraint.Hard constraint is not allow the constraint of violating.Soft-constraint is can violate if desired so that produce the constraint of the feasible solution of optimization problem.Selectively, if desired, preferably allow order that soft-constraint violated so that obtain feasible solution, this order also can be designated.

Another aspect of the present invention comprises whether the well flow of selectively specifying some well is (step 150) of being correlated with.For example, the well with similar fluid behaviour (such as output gas oil ratio (GOR) or water-oil factor (WOR)) can be relative to each other.The association of producing flow between the well will guarantee that production (or injection) flow between these wells does not vibrate arbitrarily between time step.

Then, form equation of constraint (step 160) by the well flow of producing constraint and be associated.The hard constraint equation is set up and to be used for those and not to allow the constraint of violating.Formation is corresponding to the soft-constraint equation of soft-constraint, and it comprises constraint violation punishment.Even when violating soft-constraint so that optimize when producing feasible solution, constraint violation punishment just allows to satisfy the soft-constraint equation.The generation of this group equation of constraint will be discussed in more detail below.

Create objective function in step 170, this objective function is sought the optimization aim of producing such as from the oil of oil gas field 50.Ideally, objective function comprises the composition flow rate of well and the constraint violation punishment related with the soft-constraint equation.The weighting scale factor can be associated with the soft-constraint punishment in the objective function.Appropriately weighted these weighting scale factors, then the order that can be violated with relevant soft-constraint comes this order of priorization.Come optimization aim function (step 180) by optimizing subroutine (optimizer device) then) to produce the optimized distribution of the fluid flow among the well.The optimizer device uses composition flow rate equation that calculates in step 130 and the equation of constraint of setting up in step 160 to come the optimization aim function.

Then, the fluid flow of optimization and can among well and reservoir, distribute (step 190) by other fluid flow characteristics that optimizer device (such as constraint violation punishment) is determined.Then, the flow of these optimizations and characteristic can be used as in the ensuing iteration time step-length in reservoir simulation initially/boundary condition (step 200).Repeating step 120-200 up to having passed through the satisfied time cycle, stops simulation through many time steps so that the oil gas field large-scale production of increase to be provided then then.Now above-mentioned steps will be described in more detail.

B. the foundation of objective function and equation of constraint

1. equation of constraint system

Linear programming (LP) system is one group of linear equation and linear restriction.Mixed integer programming (MIP) system is one group of linearity or nonlinear equation and constraint.In the present invention, when needs were found the solution one group of nonlinear equation being represented by piecewise linear function or retrained the target of optimizing with acquisition, preferred MIP system strengthened the LP system.Use the software program package of the open source code of LP and MIP technology to be used in this exemplary embodiment so that the optimization aim function.Especially, the present invention uses name to be called the routine package of LP-Solve, and this routine package can be from network address Http:// Packages.debian.org/stable/math/lp-solveThe place obtains.Also select to utilize name to be called the selectable commercial solver of XA, it can be from California, and the Sunset software engineering joint-stock company of San Marino obtains.Those skilled in the art will recognize, the routine package of other commercial LP/MIP optimizer devices uses fluid flow and constraint condition to can be used for the optimization aim function.

Equation of constraint, composition flow rate equation and objective function are imported in the optimizer device.The optimizer device is exported the feasible solution of optimization problem then, and this optimization problem comprises that the increase of well flow distributes.Ideally, also exported the violation value that obtains the necessary any soft-constraint of feasible solution of optimization.Then, the user can carry out the violation value of suitable change with the reflection soft-constraint to the ability of producing constraint or uphole equipment.

Seek the extreme value of objective function.Simple LP system can have following form:

$\mathrm{OBJ}=\max \left\{\underset{i}{\mathrm{\Σ}}{c}_i{x}_i\right\},$ The form that submits to restraint: (1)

$\underset{i}{\mathrm{\Σ}}[a_i{x}_i-b_i]\{\≤,\≥,=\}0$

Wherein

The i=index

c _i=weighting scale factor

x _i=optimised parameter

a _i=multiplying constant and;

b _i=addition constant.

In one embodiment of the invention, master variable is the well flow.That is to say the component of the fluid of producing from well with this flow (for example oil, water gentle).Use " flow fast " method preferably to form the composition flow rate equation, should " flow fast " method will be described below.What of amount that a kind of component of transporting by well compares with the one other fluid component composition flow rate equation described.The production flow of various ingredients relative to each other can keep linear or can be non-linear in well is produced the potential scope of exporting.Ideally, the present invention be by will systematically discussing the problem into MIP, by piecewise linear function come process element or compare between the non-linear ratio convert.Produce constraint and be built to and do not allow the hard constraint violated, and/or when needs obtain to separate, be built to the soft-constraint of permission violation.Constraint can comprise destination object and restriction of production.Objective function is set up by the information that the user provides.

2. set up objective function

Usually, objective function meets mathematic(al) representation:

$\mathrm{OBJ}=\underset{i}{\mathrm{\Σ}}{w}_i\underset{j}{\mathrm{\Σ}}{q}_{\mathrm{ij}}-\underset{k}{\mathrm{\Σ}}{w}_k{\mathrm{CVP}}_k---\left(2\right)$

Wherein

OBJ=is with optimised target;

Fluid components number in the i=wellbore fluid;

w _i=be used for the i (i of well ^Th) the weighting scale factor produced of fluid components;

J=well number;

q _Ij=by j (j ^Th) i (i producing of well ^Th) flow of component;

K=violates the punishment number with production constraint constraints associated;

w _k=the k (k ^Th) the weighting scale factor of constraint violation punishment; And

CVP _k=the k (k ^Th) constraint violation punishment.

The function of exemplary goal more specifically of LP/MIP system can comprise the weighted sum of the gentle total production flow of oil, water of selected one group of well.In the present invention, objective function also can comprise constraint violation punishment variable (CVP _k) to adapt to the use of soft-constraint.Exemplary objective function can following mathematical form be represented:

$\mathrm{OBJ}=w_o\underset{i}{\mathrm{\Σ}}{q}_{\mathrm{oi}}+w_g\underset{i}{\mathrm{\Σ}}{q}_{\mathrm{gi}}+w_w\underset{i}{\mathrm{\Σ}}{q}_{\mathrm{wi}}-\underset{k}{\mathrm{\Σ}}{w}_k{\mathrm{CVP}}_k---\left(3\right)$

Wherein

OBJ=is with optimised target;

w _oThe weighting scale factor that=oil is produced;

q _Oi=by i (i ^Th) the well oil mass of producing;

w _gThe weighting scale factor that=gas is produced;

q _Gi=by i (i ^Th) the well tolerance of producing;

w _wThe weighting scale factor of=aquatic product;

q _Wi=by i (i ^Th) the well water yield of producing;

w _k=the k (k ^Th) the weighting scale factor; And

CVP _k=the k (k ^Th) constraint violation punishment.

Weighting scale factor w _iOr the well yield parameter can be specified by the user.For example, the user can specify: w _o=1.0; w _g=-0.1; And w _w=-0.2.

These weighting ratio factor pairs should be produced the maximization of flow in oil, manage to minimize the flow of G﹠W simultaneously.In this case, objective function increases by 1.0 (w for every stock tank barrels/sky (STB/D) of the oil of production _Oil=1.0) and be the gas punishment 0.2 of every million cubic feet/sky (MSCF/D) and be the water punishment 0.1 of every STB/D.In this case, the unit of objective function is the combination of STB/D and MSCF/D unit.Ideally, carry out the normalization of objective function component so that nondimensional objective function to be provided.

If useful, the unmatched another kind of optimal way of the unit in the processing target function is to use economic information.If for example the oil income is 22$/STB/D, the gas income is 3$/MSCF/D, and the cost Shi $3.5 that handles every STB/D water, then:

w _o=22.0; w _g=3.0; And w _w=-3.5.

In this case, the unit of objective function is currency ($)) and be consistent.Determine in proportion that preferably the weighting scale factor makes w _oBe 1.0, therefore, previous well yield parameter value is drawn by with 22.0 normalization:

w _o=1.0; w _g=0.136; And w _w=-0.159.

3. produce constraint

Constraint can rise the rate limit such as the well production limit, injection flow limit, gaslift based on physical restriction.Selectively, can determine that constraint is to satisfy engineering parameter selection (such as the production/injection target of one group of well).Other constraints, the mode by example but be not limited to can comprise output gas oil ratio (GOR), water-oil factor (WOR), and about the constraint of the subclass of well or completion.

LP/MIP system restriction classification is become hard constraint and soft-constraint.For example, the maximum oil production that hard constraint can be applied to a pair of aboveground feasible combination is 5,000STB/D.These hard constraints are converted to following LP/MIP constraint:

$q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}\&le;\mathrm{5,000}---\left(4\right)$

$q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="2"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">PROD</figure-callout>}2}\&le;\mathrm{5,000}$

Wherein

And

4. the priorization of soft-constraint

When and and if only if there is not alternate manner to cash soft-constraint when obtaining the feasible solution of system simultaneously, soft-constraint is to allow the constraint of violating.Ideally, this constraint violation be obtain to separate necessary minimum may: constraint violation can take place when system and the limit/target are inconsistent.Consider following situation, promptly oil gas field has and comprises about the oily productive target of one group of well and the constraint of the water treatment limit, and is as described below:

Oil productive target=7,500STB/D (5)

The aquatic product limit＞5,000STB/D

There are a point in possibility and most probable in simulation, state the group well in this place and are not producing more than 5, under the situation of the water of 000STB/D, can not produce 7, the oil of 500STB/D.When well tenure of use was long or ripe, well was tending towards producing more water.In this case, the optimizer device will not reported not have and separates but will allow in the violation soft-constraint one.Preferably, constraint has been violated in the expression that will lift a flag.Can determine which retrains selected preferential violation by user and of the present invention preferred embodiment.

These oily targets and water electrode limit condition are converted to following three soft-constraint equations:

Constraint-1: $q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}+q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="2"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">PROD</figure-callout>}2}+{\mathrm{<figure-callout\; id="1"\; label="CVP"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">CVP</figure-callout>}}_1\&GreaterEqual;\mathrm{7,500}---\left(6\right)$

Constraint-2: $q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}+q_{p=\mathrm{oil}}^{w=\mathrm{PROD}2}-{\mathrm{<figure-callout\; id="2"\; label="CVP"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">CVP</figure-callout>}}_2\&le;\mathrm{7,500}$

Constraint-3: $q_{p=\mathrm{water}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}+q_{p=\mathrm{water}}^{w=\mathrm{PROD}2}-{\mathrm{<figure-callout\; id="3"\; label="CVP"\; filenames="S2005800525117E00092.png"\; state="\{\{state\}\}">CVP</figure-callout>}}_3\&le;\mathrm{5,000}$

Constraint violation punishment CVP _kVariable is attached on the objective function:

OBJ＝...-w ₁CVP ₁-w ₂CVP ₂-w ₃CVP ₃ (7)

Make its obedience:

w _k＞0 w wherein _kBe and k ^ThThe k that constraint violation punishment is associated ^ThThe weighting scale factor;

And

CVP _k〉=0 CVP wherein _kBe and k ^ThThe k that equation of constraint is associated ^ThConstraint violation punishment.

Notice that it is zero that this setting forces the CVP variable,,, promptly work as oil production and equal 7 that 500STB/D and aquatic product during 000STB/D, just can make them be equivalent to hard constraint less than 5 as long as can satisfy them because they have negative weight in objective function.

Suppose that reservoir conditions makes that the oil of 500STB/D must produce 5, the water of 100STB/D in order to produce 7.In this case, there are two kinds of options:

Reduce production in proportion and satisfy the water electrode limit, yet ignore oily target; Perhaps

Yet satisfy oily target production and limit more water than water electrode.

No matter the LP/MIP system selects to reduce in proportion to produce still to satisfy the water electrode limit, it depends on CVP _kThe coefficient of variable or weighting scale factor w _kThe capacity limit of supposing water is that absolute and oily production allows to reduce in proportion to produce to satisfy the water electrode limit.In this case, suppose w ₁=1; w ₂=1; And w ₃=2, this has higher priority corresponding to constraint-3 (the aquatic product limit) than other two constraints (oily productive target).Note weighting scale factor w ₃Than other two weighting scale factor w that are associated with oil production ₁And w ₂Bigger.When well yield is reduced when satisfying the aquatic product limit in proportion, suppose when aquatic product just in time be 5, during 000STB/D, oily Downturn in production to 7,400STB/D.In this case, CVP ₁To be necessary for non-zero to satisfy constraint 1, just in time be CVP ₁=100.In this setting, because CVP coefficient w _kParticular value, the LP/MIP system will select to reduce in proportion flow rather than produce more water.Following two kinds of situations will appear in the objective function input.

If ignore oily productive target and allow oil production to reduce in proportion to satisfy the water electrode limit, then:

$q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}+q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="2"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">PROD</figure-callout>}2}=\mathrm{7,400}---\left(8\right)$

$q_{p=\mathrm{water}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}+q_{p=\mathrm{water}}^{w=\mathrm{<figure-callout\; id="2"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">PROD</figure-callout>}2}=\mathrm{5,000}$

CVF ₁＝100 CVF ₂＝0 CVF ₃＝0

OBJ＝...-1CVF ₁-1CVF ₂-2CVF ₃＝...-100 (9)

If execute oily productive target, and allow the restriction of violation about the aquatic product limit, then:

$q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}+q_{p=\mathrm{oil}}^{w=\mathrm{<figure-callout\; id="2"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">PROD</figure-callout>}2}=\mathrm{7,500}---\left(10\right)$

$q_{p=\mathrm{water}}^{w=\mathrm{<figure-callout\; id="1"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">PROD</figure-callout>}1}+q_{p=\mathrm{water}}^{w=\mathrm{<figure-callout\; id="2"\; label="PROD"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">PROD</figure-callout>}2}=\mathrm{5,100}$

CVF ₁＝0 CVF ₂＝0 CVF ₃＝100

OBJ＝...-1CVF ₁-1CVF ₂-2CVF ₃＝...-200 (11)

Because all other situations are identical, reduce flow in proportion and cause higher target function value (+100), LP/MIP optimizer device will preferably reduce flow in proportion.Identical method can be used for handling n soft-constraint and they is imported with the violation priority order of wanting.

If the order of the soft-constraint that will be violated is not specified and is kept not priorization, then all weighting scale factor w _kEquate, and do not provide the priority of the constraint that preferential permission violated.In this case, w ₁=w ₂=w ₃=1.Selectively, can provide first soft-constraint is minimum priority, and providing second soft-constraint is higher a little priority, and providing the 3rd soft-constraint is the highest priority.In exemplary embodiment of the present invention, the weighting scale factor w that then provides _iValue corresponding to 10 * 10 ^p, wherein p is the priority order that soft-constraint can be violated.For example,

w ₁＝10×10 ¹ w ₂＝10×10 ² w ₃＝10×10 ³

The universal equation of objective function is:

OBJ＝...-w ₁CVP ₁-w ₂CVP ₂-w ₃CVP ₃ (12)

Then having weighting scale factor objective function becomes:

OBJ＝...-10×10 ¹CVF ₁-10×10 ²CVF ₂-10×10 ³CVF ₃ (13)

Preferably, these coefficients by normalization to be given in the value between 0 and 1.Normalization is based in part on the potential scope of constraint violation punishment.

Retrain 10＜=CVP _Normi＜=1

CVP _normi＝(CVP-CVP _min)/(CVP _max-CVP _min) (14)

Perhaps, because CVP _MinAlways zero:

w _i＝10×10 ^p/(CVP _maxi) (15)

Optimize CVP _kParameter is together with other parameters of optimization system (production/injection flow).Because CVP _kAny on the occasion of applying punishment by objective function, system manages to keep CVP _kValue is zero.And if only if does not exist other modes to obtain feasible solution, CVP _kJust obtain on the occasion of.

If note not having and optimizing inconsistent target, all CVP variablees will will be equivalent to hard constraint for zero and soft-constraint.

The operator that uses with soft-constraint will be converted into following LP/MIP equation:

Operator	The soft-constraint standard	?	Operator	The LP/MIP equation
					＞	WATPR＞5,000	Become	≤	q w≤5,000
＜	GASPR＜10,000	Become	≥	q g≥5,000
					＝	OILPR＝7500	Become	≤ and 〉=	q o≤ 7,500 and q o≥7,500

Notice that (=) operator is the target operator, and if the standard on the left side be not equal to the standard on the right, (thereby causing action) will satisfy condition.

5. related well yield

The LP/MIP system is tight accurate, thereby does not have the preferential notion of physics of variable, equation and constraint.So, in some cases, although LP/MIP result sound accurately, may produce complete impracticable sensation.Make all wells when having insignificant difference in nature when LP/MIP optimizer device determines the only well in one group of well of throttling significantly, this thing happens for possibility.This may cause the vibration of the big flow of the single well between time step.In order to prevent this situation, the invention provides options: the well yield with well of close characteristic is correlated with.

Should be correlated with if determine well yield,, also set up the equation of constraint relevant with some well flow except existing equation of constraint.For example, if having the well of the fluid behaviour in preset range each other, such as output gas oil ratio (GOR) and/or water-oil factor (WOR), then the flow of these wells can be correlated with.Similar in appearance to above-mentioned soft-constraint equation, the equation that these flows are relevant can have the weighting scale factor, and this weighting scale factor can be closer to each other and be comprised constraint violation punishment.

Referring now to Fig. 4 A, for example, provided the flow (q of well with maximum GOR ₁), the flow (q of phase closing well ₂) be allowed to be in the shaded area.This can realize by descending in the column constraint adding system:

$q_2-\frac{q_{2f}}{q_{1f}}\&CenterDot;q_1-a-\mathrm{RVP}\&le;0---\left(16\right)$

$q_2-\frac{q_{2f}}{q_{1f}}\&CenterDot;q_1-\mathrm{RVP}\&GreaterEqual;0$

Wherein:

q ₁, q ₂=the flow that is relative to each other

q _1f, q _2fThe maximum possible value of=flow

The RVP=flow is violated punishment

A=determines the value of relevant " severity "

All RVP are added in the objective function with negative weight:

$\mathrm{OBJ}=...-\underset{i}{\mathrm{\&Sigma;}}{w}_i{\mathrm{RVP}}_i---\left(17\right)$

Wherein: in this special example, select w _iBe-10.

Providing a is:

$a=f\left(\frac{{\mathrm{GOR}}_1-{\mathrm{<figure-callout\; id="2"\; label="GOR"\; filenames="S2005800525117E00011.png,S2005800525117E00071.png"\; state="\{\{state\}\}">GOR</figure-callout>}}_2}{{\mathrm{GOR}}_1}\right)q_{2}f$

This shows works as ${\mathrm{<figure-callout\; id="1"\; label="q"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">q</figure-callout>}}_1={\mathrm{<figure-callout\; id="1"\; label="q"\; filenames="S2005800525117E00011.png,S2005800525117E00042.png"\; state="\{\{state\}\}">q</figure-callout>}}_1^{*}$ The time, q ₂Need be at scope [q _2min ^*, q _2max ^*] in.Function f is the simple linear function shown in Fig. 4 B.

The present invention allows the user to change threshold value t, however t=1.0 should be able in most of the cases work, in this setting, provide t=1.0, GOR ₂=0.0 well will be incoherent, and will have the scale factor of an independent flow, although work as GOR for other extreme values ₂=GOR ₁The time, the shaded area among Fig. 4 A will be broken down into the line shown in Fig. 4 B, and will force second well to have identical scale factor with well 1.

It is by flow in the one group of well that converts in proportion with the identical factor that another kind makes the relevant mode of flow.For example, the production flow of all production devices of the injection flow of all injectors of the first injection well group and the well first production group can be relevant.This in this case relevant be not based on GOR or WOR; During the relevant flow that only means when the well that converts in proportion with the factor, other well in relevant the group will convert in proportion by same factors.

For example, if the well in the first production group need reduce by half production (perhaps to satisfy other constraint), then all wells in the first production group reduce by half production.Ideally, this relevant default value has in the LP/MIP system than specifying the little weight of constraint.This means in order to satisfy constraint, can break discharge relation.Parameter can be used for determining the relative weighting of constraint and the discharge relation in the LP/MIP system.These coefficients more little (negative value), these coefficients will influence system more.

C. the generation of flow curve and composition flow rate equation

1. quick discharge method

Following " flow fast " method is preferred for forming the composition flow rate curve and the equation of fluid stream.Flow curve is narrated a kind of production of component and how the production of another kind of component compares.For example, when aboveground flow controller or valve are opened, oil, the gentle production of water will increase usually.Being increased in total fluid production scope between any two kinds of components can be linear or non-linear.Referring again to Fig. 3 A and 3B, the production of shown G﹠O totally is linear and the production of water and oil is non-linear generally.Flow curve is produced by a series of data points.Use the Newton-Raphson iterative program to represent by " x " mark together with the data point of the subdivision generation of reservoir model.The data point of representing with " square " mark is to use " flow fast " method to produce.Notice that two kinds of methods provide similar result.Yet the counting yield of " flow fast " method is much higher.

The fact of discharge method utilization is at the set time point fast, the common and linear ratio of pressure drop of the production of single well completion unit.Pressure drop is pressure in the well well completion unit and the pressure reduction between the adjacent reservoir units.Corresponding produce and implant operation during be that this pressure reduction orders about fluid and enters and flow out well completion unit.Use the many different pressure drop of each well section to produce one group of data point.Then, construct the piecewise linear function of the most suitable these points ideally.Produce the composition flow rate equation by this piecewise linear function that optimised sequencer is used then.

The total composition flow rate curve of oil-water is piecewise linear, and it is not a linear function.Fig. 5 A-D has shown the flow of the single well completion unit of four different total growths that are used for well.The pressure traverse that has also shown the well unit of the different production of reservoir flow with these.The situation of Fig. 5 A-D explanation is: such as when the well top of well throttling valve is closed, oil production is taken place continue to reduce.Attention is along with oil yield reduces, and the aquatic products amount reduces up to producing water hardly.

When the production flow reduces, the borehole pressure section of well will increase.Suppose preset time the step-length reservoir pressure traverse keep constant.This cause pressure drop in the well along with the borehole pressure section to the increase of reservoir pressure section and reduce.Note because the influence of pressure head/gravity is big with the completion of the superficial degree of depth than the pressure of dark completion.Therefore, pressure drop is lower in the darker degree of depth, at the bigger water of this degree of depth place density below the less oil and gas-bearing formation of density.

The present invention utilizes linear ratio's single well completion that converts in proportion.The flow sum that the total production flow that comes the component p (i.e. oil, water or gas) of artesian well is its mobile completion:

$q_{\mathrm{pT}}=\underset{i=1}{\overset{n_{\mathrm{comp}}}{\mathrm{\&Sigma;}}}{q}_{\mathrm{pi}}---\left(19\right)$

Wherein

q _PT=come the total amount of the stream of artesian well;

n _CompWell completion unit number in the=well; And

q _Pi=from i (i ^Th) flow of component stream of well.

Standard flow in each component at each single completion place extracts from the reservoir simulation operation with special time step and well production level.Suppose completion flow and the linear ratio of pressure drop in each single well completion unit of fixed point of time point.Thereby if pressure drop reduces with a certain quantity c in the well, the flow of single completion will correspondingly reduce in proportion, and will be to the total flow that makes new advances:

$q_{\mathrm{pT}}^{*}=\underset{i=1}{\overset{n_{\mathrm{comp}}}{\mathrm{\&Sigma;}}}\frac{\mathrm{\&Delta;}{P}_i-c}{\mathrm{\&Delta;}{P}_i}{q}_{\mathrm{pi}}---\left(20\right)$

Wherein

Reducing of c=pressure drop;

n _CompWell completion unit number in the=well;

Δ P _i=the i (i ^Th) initial drop in the well completion unit; And

q _Pi=from i (i _Th) well mutually stream amount;

Thereby, provide and the oily flow of well need be reduced to q from q ^*The time the side-play amount c of pressure:

$c=\frac{q_{\mathrm{oT}}-q_{\mathrm{oT}}^{*}}{\underset{i=1}{\overset{n_{\mathrm{comp}}}{\mathrm{\&Sigma;}}}\frac{q_{\mathrm{oi}}}{\mathrm{\&Delta;}{P}_i}}---\left(21\right)$

The parallel offset in the borehole pressure section has been indicated in this pressure skew, as shown in Figure 6.C as calculated, equation 20 can be used for calculating the well yield that other components of well flow.Identical program also can be used for injecting flow.As before considering, repeat this program and can produce the data point of many component stream and can form curve about Fig. 3 A and 3B.

2. the structure of piecewise linear function

Produce piecewise linear function, it preferably represents these data points that are used for each well by " flow fast " method generation.

Piecewise linear function comprises many line segments and breakpoint.Use least square fitting by " flow fast " quantity and the position of data set that method produced to select breakpoint ideally.In this exemplary embodiment, Levenberg-Marquardt least square fitting method is preferred for location break point.The data point of using by the optimizer device that those skilled in the art will recognize that other curves or equation generation technique can be used for representing producing.

Referring now to Fig. 7 A and 7B, for given line segment k, the coordinate that provides the line segment end points is:

(a _2k-1, a _2k) and (a _2k+1, a _2k+2) (22)

The difference quotient y that least square method (such as Levenberg-Marquardt) needs this function, determine according to parameter:

A. these difference quotients are:

$\frac{\&PartialD;y}{\&PartialD;a_{2k-1}}=(a_{2k+2}-a_{2k})\frac{-a_{2k+1}+x}{{(a_{2k+1}-a_{2k-1})}^2}---\left(23\right)$

$\frac{\&PartialD;y}{\&PartialD;a_{2k}}=1-\frac{x-a_{2k-1}}{a_{2k+1}-a_{2k-1}}$

$\frac{\&PartialD;y}{\&PartialD;a_{2k+1}}=\frac{(x-a_{2k-1})(a_{2k+2}-a_{2k})}{{(a_{2k+1}-a_{2k-1})}^2}$

$\frac{\&PartialD;y}{{\&PartialD;a}_{2k+2}}=\frac{(x-a_{2k-1})}{a_{2k+1}-a_{2k-1}}$

In a preferred embodiment, the breakpoint of right quantity and their optimum position have been determined ideally.Algorithm as shown in Figure 8 is used to select breakpoint number.

First step is by linear function (for example single line segment, two end points) beginning, so i=2.Calculate this linear function χ _i ²Be χ ₂ ²Then breakpoint is added in the linear function, it is become have two sections and three end points (i=i+1, that is, and piecewise linear function i=3).The breakpoint coordinate is optimised for minimum χ _i ²If away from initial match, then add new breakpoint by the match that improves more than one the f factor, it is not obvious up to improving to repeat this process.Have only when improving this match by coefficient f, this algorithm keeps adding more multibreak.

Dependence has the line segment of bigger quantity, can carry out better match by reducing the f value.The common practicality of this method is very strong.Can check to guarantee that breakpoint always is in (first quartile) in the feasible zone.This can fall into separating of infeasible region by punishment (P) guarantees, as shown in Figure 9.

3. piecewise linear function is merged into linear programming

Piecewise linear curve is merged into other continuous parameters and some constraints that LP is provided with needs to introduce binary.Following is a set of equations and the variable that needs interpolation:

Breakpoint:

(x _bi，y _bi) i＝1，2...，n

Be used for replacing the flow item:

q＝z ₁y _b1+z ₂y _b2+...+z _ny _bn

Add constraint:

z ₁≤y ₁ z ₂≤y ₁+y ₂ z ₃≤y ₂+y ₃?...?z _n-1≤y _n-2+y _n-1 z _n≤y _n-1

y ₁+y ₂+...+y _n-1＝1 (25)

z ₁+z ₂+...+z _n＝1

q ₁＝z ₁x _b1+z ₂x _b2+...+z _nx _bn

y ₁∈{0，1} i＝1，2...，n-1

z ₁≥0 i＝1，2...，n

Q is the subordinate flow herein, q ₁It is Control Flow.To prove that now this set why causes having the correct behavior of two sections simple piecewise linear function.Assumed function as shown in Figure 10.With the functional value of determining at the x=15 place.With this problem corresponding formulas will for:

f(x)＝z ₁0+z ₂3+z ₃9 (26)

x＝z ₁0+z ₂20+z ₃30

z ₁≤y ₁ z ₂≤y ₁+y ₂ z ₃≤y ₂

y ₁+y ₂＝1

z ₁+z ₂+z ₃＝1

y _i∈{0，1} i＝1，2

z _i≥0 i＝1，2，3

The y of binary represents the line segment that x belongs to.In this case.y ₁Should be 1 and y ₂Should be 0.At first check and see y ₂Whether once be 1.If y ₂Once be 1, y then ₁Be necessary for 0, this means z ₁Be 0 and z ₂And z ₃It is non-zero.Yet, if z ₂And z ₃Be non-zero, may satisfy the x=15 of second equation never, then y ₂Can not be 1.If thereby y ₁Be 1, then find the solution z and draw:

z ₁＝0.25

f(x)＝2.25

z ₂＝0.75

The merging of the variable in the combination of described equation and the equation 24 forces LP/MIP optimizer device to realize the composition flow rate curve.

Though described above stated specification of the present invention with and some relevant preferred embodiment, and many particulars are provided in order to explain, yet the present invention allows and changes and some other particulars described herein can appropriate change and can not depart from ultimate principle of the present invention that this is conspicuous for a person skilled in the art.

Claims (15)

1. method that is used to increase the flow distribution of fluid among a plurality of wells, described a plurality of wells are communicated with at least one subsurface reservoir fluid, and described method comprises:

(a) simulation at least one subsurface reservoir and with a plurality of wells that described at least one subsurface reservoir fluid is communicated with in comprise the flow of multi-component fluid;

(b) selection comprises the production constraint of at least one soft-constraint and at least one hard constraint, and wherein said at least one soft-constraint is violable, and described at least one hard constraint is observed;

(c) produce system equation and equation of constraint, described system equation comprises the composition flow rate equation corresponding to the fluid stream that simulated in the well, the characteristic that comprises the fluid stream in more at least two wells, if described characteristic is in preset range each other, then the fluid flow of described at least two wells is associated together, makes described at least two wells will have relevant dispense flow rate by in system equation, producing the relevant equation of flow; Described equation of constraint comprises at least one the soft-constraint equation that is associated with described at least one soft-constraint, and described at least one soft-constraint equation comprises the constraint violation punishment (CVP) that allows described at least one soft-constraint equation to satisfy soft-constraint;

(d) produce corresponding to the stream of the fluid in the well and corresponding to the objective function of constraint violation punishment;

(e) utilize optimizer device and described system equation to come the optimization aim function to distribute with the increase of determining the fluid flow among a plurality of wells, wherein, if desired, can violate energy actual the separating of adopting of described at least one soft-constraint to obtain to optimize, and, violate described at least one hard constraint and will cause this optimization can not adopt in practice; And

(f) control fluid and flow into and flow out described a plurality of well by regulating well control system device, thereby as in step (e) determined among a plurality of wells the distributing fluids flow.

2. the method for claim 1, wherein

Described production constraint comprises violable a plurality of soft-constraint;

Described system equation comprises a plurality of soft-constraint equations corresponding to described a plurality of soft-constraints, and each described soft-constraint equation comprises the corresponding constraint violation punishment (CVP) that allows the satisfied soft-constraint separately of soft-constraint equation; And

Described objective function is punished corresponding to the stream of the fluid in the well and corresponding to constraint violation;

Wherein, if desired, can violate energy actual the separating of adopting of described soft-constraint to obtain to optimize.

3. method as claimed in claim 2, wherein:

Come the described soft-constraint of priorization according to the difficulty of the soft-constraint that will be violated.

4. method as claimed in claim 3, wherein:

In objective function, the punishment of the constraint violation of weighting scale factor and corresponding soft-constraint equation is associated, and is weighted according to the priorization with the soft-constraint of separately soft-constraint dependence among equations connection, so that make the more difficult violation of the soft-constraint of higher priority.

5. the method for claim 1, wherein:

Described objective function meets mathematic(al) representation:

$\mathrm{OBJ}=\underset{i}{\mathrm{\&Sigma;}}{w}_i\underset{j}{\mathrm{\&Sigma;}}{q}_{\mathrm{ij}}-\underset{k}{\mathrm{\&Sigma;}}{w}_k{\mathrm{CVP}}_k$

Wherein

OBJ=is with optimised target;

Fluid components number in the i=fluid;

w _iThe weighting scale factor of i fluid production the in=well;

J=well number;

q _IjThe flow of=i the component of producing by j well;

K=and soft-constraint constraints associated are violated the punishment number;

w _kThe weighting scale factor of=the k constraint violation punishment; And

CVP _k=the k constraint violation punishment.

6. method as claimed in claim 2, wherein:

Described objective function meets mathematic(al) representation:

$\mathrm{OBJ}=\underset{i}{\mathrm{\&Sigma;}}{w}_i\underset{j}{\mathrm{\&Sigma;}}{q}_{\mathrm{ij}}-\underset{k}{\mathrm{\&Sigma;}}{w}_k{\mathrm{CVP}}_k$

Wherein

OBJ=is with optimised target;

Fluid components number in the i=fluid;

w _iThe weighting scale factor that i fluid components the in=well produced;

J=well number;

q _IjThe flow of=i the component of producing by j well;

K=and soft-constraint constraints associated are violated the punishment number;

w _kThe weighting scale factor of=the k constraint violation punishment; And

CVP _k=the k constraint violation punishment.

7. method as claimed in claim 6, wherein:

Come the described soft-constraint of priorization according to the difficulty of the soft-constraint that will be violated; And

Constraint violation punishment CVP with soft-constraint equation separately _kThe described weighting scale factor w that is associated _kPriorization according to described soft-constraint is weighted, so that make the more difficult violation of the soft-constraint of higher priority.

8. the method for claim 1, wherein:

Described well comprises a plurality of well completion unit, and described at least one subsurface reservoir comprises a plurality of reservoir units that are communicated with described well completion unit fluid;

The step of simulation fluid flow comprises to be determined in the reservoir units and the pressure in the well completion unit, and comprises owing to the pressure drop between reservoir units and well completion unit, and determines the fluid flow of the respective components in the well completion unit; And

Described composition flow rate equation is produced by the data point of composition flow rate, based on determined component flow flow in the fluid flow field simulation with respect to the pressure drop of the change between reservoir units and well completion unit, described data point by by than rate conversion and summation the component flow stream in the well completion unit of each well set up.

9. method as claimed in claim 8, wherein:

The composition flow rate data point utilizes following mathematic(al) representation to produce:

$q_{\mathrm{pT}}^{*}=\underset{i=1}{\overset{n_{\mathrm{comp}}}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&Delta;P}}_i-c}{{\mathrm{\&Delta;P}}_i}{q}_{\mathrm{pi}}$

Wherein:

=come the new total flow of artesian well;

n _CompWell completion unit number in the=specific well;

Δ P _i=i ^ThThe initial drop of well completion unit;

C=is from the change in pressure drop of the pressure drop of the primary simulation of well completion unit; And

q _Pi=from the primary simulation flow of the component stream of i well completion unit;

10. method that is used to increase the flow distribution of fluid among a plurality of wells, described a plurality of wells are communicated with at least one subsurface reservoir fluid, and described method comprises:

(a) flow that comprise multi-component fluid of simulation in a plurality of wells and at least one subsurface reservoir, described well comprises a plurality of well completion unit, described at least one subsurface reservoir comprises a plurality of reservoir units that are communicated with described well completion unit fluid; Determine in the reservoir units and well completion unit in pressure; And, determine the flow of the respective components in the well completion unit owing to the pressure drop between reservoir units and well completion unit;

(b) selection comprises the production constraint of at least one soft-constraint and at least one hard constraint, and wherein said at least one soft-constraint is violable, and described at least one hard constraint is observed;

(c) based on determined component flow flow in step (a) and the pressure drop that changes between with respect to reservoir units and well completion unit, by the stream of the component flow in the described well completion unit is pressed than rate conversion and summation, in the scope of fluid flow, produce the data point of the composition flow rate of well;

(d) based on the data point of each well, produce the composition flow rate equation of described well, the characteristic that comprises the fluid stream in more at least two wells, if described characteristic is in preset range each other, then the fluid flow of described at least two wells is associated together, makes described at least two wells will have relevant dispense flow rate by in the composition flow rate equation, producing the relevant equation of flow;

(e) produce corresponding to the equation of constraint of producing constraint, equation of constraint comprises at least one the soft-constraint equation that is associated with described at least one soft-constraint, and described at least one soft-constraint equation comprises the constraint violation punishment (CVP) that allows described at least one soft-constraint equation to satisfy soft-constraint;

(f) produce the objective function that flows corresponding to the fluid in the well;

(g) utilize optimizer device and described equation of constraint and composition flow rate equation to come the optimization aim function, distribute with the increase of determining the fluid flow among a plurality of wells, wherein, if desired, can violate energy actual the separating of adopting of described at least one soft-constraint to obtain to optimize, and, violate described at least one hard constraint and will cause this optimization can not adopt in practice; And

(h) control fluid and flow into and flow out described a plurality of well by regulating well control system device, thus according in step (g) determine and among a plurality of wells the distributing fluids flow.

11. method as claimed in claim 10 also comprises:

Data point by each well produces piecewise linear function; And

Produce the composition flow rate equation by described piecewise linear function.

12. method as claimed in claim 11, wherein

Described composition flow rate equation comprises binary variable so that describe piecewise linear function; And

Optimization step comprises use mixed integer programming.

13. method as claimed in claim 10, wherein

Described optimizer device utilizes at least a in linear programming and the mixed integer programming in the optimization aim function.

14. method as claimed in claim 13, wherein

Described optimizer device utilizes mixed integer programming in the optimization aim function.

15. method as claimed in claim 10, wherein

The composition flow rate data point utilizes following mathematic(al) representation to construct:

$q_{\mathrm{pT}}^{*}=\underset{i=1}{\overset{n_{\mathrm{comp}}}{\mathrm{\&Sigma;}}}\frac{{\mathrm{\&Delta;P}}_i-c}{{\mathrm{\&Delta;P}}_i}{q}_{\mathrm{pi}}$

Wherein:

=from the new total flow of well;

n _CompWell completion unit number in the=certain well; And

Δ P _iThe initial drop of=the i well completion unit;

C=is from the change in pressure drop of the pressure drop of the primary simulation of well completion unit; And

q _Pi=from the primary simulation flow of the component stream of i well completion unit.

Images