CN105117809A - Method for optimizing well drilling sequence of offshore cluster well group - Google Patents

Abstract

The present invention discloses a method for optimizing a well drilling sequence of an offshore cluster well group. The method includes a well drilling sequence optimizing index, an optimizing model and a genetic algorithm solution, specially includes the steps as follows: presetting a well trajectory design scheme (containing wellhead groove parameters) of the whole well group; performing interpolation calculation (separation distance 1m), coordinate transformation, and minimum range sweep analysis for all designing trajectories; giving a measure error to all designing trajectories so as to calculate error ellipsoid parameters corresponding to all interpolating points; performing a well colliding risk analysis for the whole well group (calculating a well separation coefficient); building a well drilling sequence optimal object matrix and a mathematical programming model; solving the model by utilizing a genetic algorithm; and analyzing data by synthesis (giving consideration to a batch drilling requirement). The optimized well drilling sequence facilitates reducing the well colliding risk and an anti-colliding workload, facilitates saving a well drilling time and a well drilling cost by utilizing the batch well drilling method, and satisfies the requirement for optimizing the drilling sequence of the offshore large-scale cluster well group.

Description

Marine collecting well group drilling well sequence optimizing method

Technical field

The present invention relates to offshore oil drilling engineering field, particularly relate to marine large-scale collecting well group drilling well sequence optimizing method.

Background technology

Marine collecting well group treats that drilling well is more and well head spacing is less, Drilling optimization order to exploitation and drilling well all significant, not only contribute to implementing reserves and reservoir, reduce development risk, also contribute to anti-collision and batch drilling operation, save platform holding time.When developing department does not specify drilling well order, drilling well department Main Basis anti-collision and batch drilling well require preferred drilling well order.Existing document only gives drilling well order optimum principle, not yet provides corresponding optimizing index and Optimized model.Such as, the well that when considering that anti-collision requires, the peripheral horizontal shift of usual subdrilling is large, whipstock point is shallow, rear brill internal drainage prosposition moves little, that whipstock point is dark well; Consider that batch drilling well adopts the clinodiagonal drilling well order etc. of jumping " day " word or sphere of movements for the elephants usually when requiring.From permutation and combination in theory, the drilling well of n mouth well order have n! Kind.Drilling well order has 720 kinds, 40320 kinds, 3628800 kinds respectively such as, when total well number is respectively 6 mouthfuls, 8 mouthfuls, 10 mouthfuls! Obviously, when (being no less than 10 mouthfuls of wells) when drilling well is more, only relying on above-mentioned optimization principles can only find preferably " feasible solution ", only just can find " optimum solution " by mathematic optimal model and Efficient Solution method.

Mainly there is following shortcoming in traditional marine collecting well group drilling well sequence optimizing method: (1) only has anti-collision and batch drilling optimization principle, entirely by rule of thumb with artificial selection drilling well order; (2) lack drilling well sequential optimization evaluation index, be difficult to carry out evaluation with preferred to some " feasible solution ", be difficult to take into account anti-collision and batch drilling well requirement; (3) mathematic optimal model and Efficient Solution method is lacked, when can not find " optimum solution " until drilling well more (being not less than 10 mouthfuls of wells) time.

Summary of the invention

The present invention is directed to existing marine cluster wells drilling well sequence optimizing method Problems existing and deficiency, headed by anti-collision, want target, take into account batch drilling well requirement, give complete, practical marine collecting well group drilling well sequence optimizing method.

The technology used in the present invention solution is:

A kind of marine collecting well group drilling well sequence optimizing method, comprises the following steps:

(1) the Hole clean scheme of given whole well group;

(2) adopt well interpolation method, interpolation calculation is carried out to whole designed path, and replaces original designed path with the track after interpolation processing;

(3) whole designed path is transformed into in certain reference point coordinate system that is initial point;

(4) according to anti-collision scan method, minimum distance scanning analysis is carried out to whole designed path;

(5) whole designed path is composed with measuring error, according to well track error analysis method, calculate the error ellipsoid parameter that whole interpolation point is corresponding;

(6) according to well separation coefficient computing method, the whole interpolation point of this well designed path well separation coefficient corresponding with all the other each wells is calculated by mouth well;

(7) all interpolation point numberings well separation coefficient on each designed path being less than set-point are stored in S set respectively _ijin, i, j=1,2 ..., n, finally arranges out objective matrix S, sees formula (1);

$S=\left[\begin{array}{cccccc}\mathrm{\φ}& S_{12}& ...& S_{1j}& ...& S_{1n}\\ S_{21}& \mathrm{\φ}& ...& S_{2j}& ...& S_{2n}\\ .& .& & .& & .\\ .& .& & .& & .\\ .& .& & .& & .\\ S_{i1}& S_{i2}& ...& S_{ij}& ...& S_{in}\\ .& .& & .& & .\\ .& .& & .& & .\\ .& .& & .& & .\\ S_{n1}& S_{n2}& ...& S_{nj}& ...& \mathrm{\φ}\end{array}\right],(i,j=1,2,\mathrm{...},n)---\left(1\right)$

In objective matrix S, each element is a subclass, S set _ijbe with i well for stratigraphic well, during with j well for comparing well, on i well designed path, separation coefficient is less than whole node serial numbers of set-point; The S set as i=j _ij=φ, φ are empty set;

(8) be numbered from 1 to n n mouth well, replace pound sign by numbering, each drilling well order all correspondence one has the fully intermeshing of n element; If fully intermeshing P is any one drilling well order, i.e. P=[p ₁, p ₂..., p _k... p _n] (k=1,2 ..., n; 1≤p _k≤ n); Set up drilling well sequential optimization model so that the anti-collision well section total length of whole well group is minimum for optimization aim, see formula (2);

$L=min\underset{k=2}{\overset{n}{\Σ}}card\left(\underset{m=1}{\overset{k-1}{\∪}}{S}_{p_k{p}_m}\right)\·\mathrm{\Δ}L---\left(2\right)$

In formula, L is the anti-collision well section total length of the whole well group that drilling well order P is corresponding, m; represent k-1 S set _pkpmunion, identical element not double counting; The number of element in card (*) × expression set; Δ L is borehole track interpolation point spacing, m;

(9) genetic algorithm for solving drilling well sequential optimization model is adopted;

(10) the multiple optimum solution provided for genetic algorithm and suboptimal solution, in conjunction with batch drilling well requirement, optimize the drilling well order being conducive to anti-collision and batch drilling well most.

In step (5): first calculate the error ellipse of perpendicular hole section correspondence and the circle of uncertainty of equivalence one-tenth area equation, then calculate error ellipsoid corresponding to lower curved section.

In step (9): described genetic algorithm comprises to be selected initial population and iterations, structure fitness function and optimizes genetic operator, and genetic operator comprises selection opertor, crossover operator and mutation operator; Specific as follows:

(1) initial population: N group sequence is produced at random for n mouth well individual, as initial population, get initial population N=4n;

(2) fitness function: by formula (3) structure fitness function

$f\left(i\right)=1-\frac{L\left(i\right)-L_{\min }}{L_{max}-L_{\min }+\mathrm{\Δ}L}---\left(3\right)$

In formula, f (i) is i-th individual corresponding adaptive value in initial population; L (i) is i-th individual corresponding anti-collision well section total length, m; L _maxand L _minbe respectively maximal value and the minimum value of anti-collision well section total length in initial population, m; Δ L is borehole track interpolation point spacing, m;

(3) selection opertor: initial population is selected in adoption rate system of selection;

(4) crossover operator: each drilling well order all correspondence one has the fully intermeshing of n element, adopts partial mapped crossover method, is avoid parent to intersect to produce poor filial generation, also adds crosscheck step; Crossover operator basic step is as follows:

1. according to well number n, the width w of halved belt is determined by formula (4), then the reference position of random selecting halved belt;

$w=\mathrm{int}\left(\frac{n}{5}\right)+2---\left(4\right)$

2. exchange the halved belt of parents, form original filial generation;

3. the mapping relations of two halved belts are determined;

4. according to mapping relations, the non-crossing band of original filial generation is changed, until filial generation does not conflict;

(5) mutation operator: adopt and exchange variation method, to a certain drilling well order Stochastic choice two positions, then exchanges the pound sign on these two positions mutually; In order to avoid precocious phenomenon, carry out pressure populational variation equally, and add variation checking procedure prevent variation produce poor filial generation;

(6) iterations: determine iterations by formula (5);

$C\≥50\×\mathrm{int}(\frac{n}{5}+1)---\left(5\right)$

In formula, n is collecting well group well number; C is iterations.

Advantageous Effects of the present invention is:

(1) the present invention considers the importance of marine collecting well group anti-collision problem, proposes using whole collecting well group anti-collision well section total length minimum as drilling well sequential optimization index first, can carry out quantitative evaluation with preferred to given drilling well order.

(2) the present invention gives marine collecting well group drilling well sequential optimization model and genetic algorithm for solving method first, with in the past by rule of thumb with artificial selection drilling well order compared, substantially increase drilling well sequential optimization quality (can optimum solution be found) and efficiency.

(3) the present invention is to meet anti-collision for the most important condition, take into account batch drilling well requirement, the marine cluster wells drilling well sequence optimizing method provided advantageously, contributes to reducing anti-collision difficulty of construction and workload, also contribute to batch drilling mode, save rig time and drilling cost.

(4) the present invention is when carrying out well track error analysis, propose first to calculate the error ellipse of perpendicular hole section correspondence and the circle of uncertainty of equivalence one-tenth area equation, calculate the error ellipsoid that lower curved section is corresponding again, well is handed over and touches risk evaluation results and drilling well sequential optimization result is more reliable.

(5) the marine cluster wells drilling well sequential optimization model that the present invention sets up there will be usually separates phenomenon more; can rerun about 10 times to genetic algorithm; and all reset initial population at every turn; finally export All Optimal Solutions and multiple suboptimal solution, require to do further screening so that Drilling Design personnel combine batch drilling well.

Accompanying drawing explanation

Below in conjunction with accompanying drawing and embodiment, the invention will be further described:

Fig. 1 is the process flow diagram of the present invention's marine collecting well group drilling well sequence optimizing method;

Fig. 2 is the process flow diagram that in the present invention, well track error analysis is corresponding;

Fig. 3 is the process flow diagram that in the present invention, genetic algorithm for solving process is corresponding;

Fig. 4 is the crossover operator process flow diagram of genetic algorithm in the present invention;

Fig. 5 is the well head groove Rankine-Hugoniot relations figure of the marine collecting well group of THE WESTERN SOUTH CHINA SEA.

Embodiment

By reference to the accompanying drawings, a kind of marine collecting well group drilling well sequence optimizing method, comprises the following steps:

(1) the Hole clean scheme of given whole well group, containing whole designed path and well head groove parameter in scheme.

(2) well interpolation analysis: adopt well interpolation method, interpolation calculation (getting interpolation point spacing 1.0m) is carried out to whole designed path, and replaces original designed path with the track after interpolation processing.

(3) coordinate conversion: whole designed path is transformed into certain reference point (Platform center, or certain well head groove center) as in the coordinate system of initial point.Formula of Coordinate System Transformation is shown in formula (6):

ΔN _i＝X _i0-X ₀

ΔE _i＝Y _i0-Y ₀

ΔD _i＝Z _i0-Z ₀(6)

${\hat{N}}_{ij}=N_{ij}+{\mathrm{\ΔN}}_i$

${\hat{E}}_{ij}=E_{ij}+{\mathrm{\ΔE}}_i$

${\hat{D}}_{ij}=D_{ij}+{\mathrm{\ΔD}}_i$

In above formula, X ₀, Y ₀, Z ₀for the terrestrial coordinate that reference point is corresponding, be followed successively by northern southern coordinate, thing coordinate, height above sea level; X _i0, Y _i0, Z _i0be the terrestrial coordinate that i-th mouthful of well well head groove center is corresponding, be followed successively by northern southern coordinate, thing coordinate, height above sea level; N _ij, E _ij, D _ijbe the corresponding local coordinate (note: with this well well head groove center for initial point) of a jth interpolation point on i-th mouthful of well designed path, be followed successively by N coordinate, E coordinate, vertical depth; be the world coordinates that on i-th mouthful of well designed path, a jth interpolation point is corresponding, be followed successively by N coordinate, E coordinate, vertical depth.

(4) anti-collision scan analysis: according to anti-collision scan method, carries out minimum distance scanning analysis to whole designed path, calculates interpolation point minimum distance.

(5) well track error analysis: whole designed path is composed with measuring error, according to well track error analysis method, carries out well track error analysis by drilling trajectory, calculate the error ellipsoid parameter that whole interpolation point is corresponding.Consider that whipstock point above perpendicular hole section position angle is uncertain, error ellipsoid spatial attitude is also uncertain, if artificial designated parties parallactic angle is just equivalent to specify error ellipsoid spatial attitude, can affects well friendship and touch analysis result.First calculate the error ellipse of whipstock point above perpendicular hole section correspondence in the present invention and the circle of uncertainty of equivalence one-tenth area equation, then calculate error ellipsoid corresponding to whipstock point lower curved section, concrete steps as shown in Figure 2.First given whole designed path (result interpolation processing), given well track measuring error (comprising hole drift angle, position angle and well depth survey error), then error ellipsoid parameter (arbitrary orientation) corresponding to whipstock point above straight well section interpolation point is calculated by well, again error ellipse corresponding for diameter segment interpolation point equivalence is become the circle of uncertainty, finally calculate error ellipsoid parameter corresponding to whipstock point following lower curved section interpolation point by well.Well friendship can be made to touch risk evaluation results for the method and drilling well sequential optimization result is more reliable.

(6) well is handed over and is touched venture analysis: according to well separation coefficient computing method, carries out well friendship and touches venture analysis, calculate the whole interpolation point of this well designed path well separation coefficient corresponding with all the other each wells by mouth well to whole well group.

(7) establishing target matrix: hand over based on well and touch venture analysis result, all interpolation point numberings well separation coefficient on each designed path being less than set-point (usually getting 1.5) are stored in S set respectively _ijand S _ji(i, j=1,2 ..., n), finally arrange out objective matrix S (n × n matrix), see formula (1);

$S=\left[\begin{array}{cccccc}\mathrm{\φ}& S_{12}& ...& S_{1j}& ...& S_{1n}\\ S_{21}& \mathrm{\φ}& ...& S_{2j}& ...& S_{2n}\\ .& .& & .& & .\\ .& .& & .& & .\\ .& .& & .& & .\\ S_{i1}& S_{i2}& ...& S_{ij}& ...& S_{in}\\ .& .& & .& & .\\ .& .& & .& & .\\ .& .& & .& & .\\ S_{n1}& S_{n2}& ...& S_{nj}& ...& \mathrm{\φ}\end{array}\right],(i,j=1,2,\mathrm{...},n)---\left(1\right)$

In objective matrix S, each element is a subclass, S set _ijand S _jinot necessarily identical.Because S set _ijbe with i well for stratigraphic well, during with j well for comparing well, on i well designed path, separation coefficient is less than whole node serial numbers of set-point; S _jibe with j well for stratigraphic well, during with i well for comparing well, on j well designed path, separation coefficient is less than whole node serial numbers of set-point; The S set as i=j _ij=φ, φ are empty set.

(8) build drilling well sequential optimization model: be numbered from 1 to n n mouth well, replace pound sign by numbering, each drilling well order all correspondence one has the fully intermeshing of n element.If fully intermeshing P is any one drilling well order, i.e. P=[p ₁, p ₂..., p _k... p _n] (k=1,2 ..., n; 1≤p _k≤ n).Set up drilling well sequential optimization model so that the anti-collision well section total length of whole well group is minimum for optimization aim, see formula (2).

$L=min\underset{k=2}{\overset{n}{\Σ}}card\left(\underset{m=1}{\overset{k-1}{\∪}}{S}_{p_k{p}_m}\right)\·\mathrm{\Δ}L---\left(2\right)$

In formula, L is the anti-collision well section total length of the whole well group that drilling well order P is corresponding, m; represent k-1 S set _pkpmunion (identical element not double counting); Card (*) represents the number of element in set; Δ L is borehole track interpolation point spacing (at equal intervals), m;

(9) genetic algorithm for solving drilling well sequential optimization model: genetic algorithm comprises to be selected initial population and iterations, structure fitness function and optimizes genetic operator (comprising selection opertor, crossover operator and mutation operator) etc.Specific as follows:

(1) initial population: N group sequence is produced at random for n mouth well individual, as initial population.According to different well number the trial result, recommend to get initial population N=4n, the likelihood ratio obtaining optimum solution is comparatively large and iteration time can not be too much.

(2) fitness function: the drilling well sequential optimization problem that formula (2) provides belongs to minimum problems, needs by formula (3) structure fitness function.

$f\left(i\right)=1-\frac{L\left(i\right)-L_{\min }}{L_{max}-L_{\min }+\mathrm{\Δ}L}---\left(3\right)$

In formula, f (i) is i-th individual corresponding adaptive value in initial population; L (i) is i-th individual corresponding anti-collision well section total length, m; L _maxand L _minbe respectively maximal value and the minimum value of anti-collision well section total length in initial population, m; Δ L is borehole track interpolation point spacing, m.

(3) selection opertor: initial population is selected in adoption rate system of selection.

(4) crossover operator: each drilling well order all correspondence one has the fully intermeshing of n element, the most applicable employing partial mapped crossover method.Intersect for avoiding parent and produce poor filial generation, also need to add crosscheck step.Crossover operator basic step is as follows:

1. according to well number n, the width w of halved belt is determined by formula (4), then the reference position of random selecting halved belt;

$w=\mathrm{int}\left(\frac{n}{5}\right)+2---\left(4\right)$

2. exchange the halved belt of parents, form original filial generation;

3. the mapping relations of two halved belts are determined;

4. according to mapping relations, the non-crossing band of original filial generation is changed, until filial generation does not conflict.

(5) mutation operator: drilling well sequential optimization problem is applicable to adopting exchanges variation method.Also namely, to a certain drilling well order Stochastic choice two positions, then the pound sign on these two positions is exchanged mutually.In order to avoid precocious phenomenon, carry out pressure populational variation equally, and add variation checking procedure prevent variation produce poor filial generation.

(6) iterations: at present main rule of thumb or calculation method determination iterations.According to the trial result, recommend to determine iterations by formula (5).

$C\≥50\×\mathrm{int}(\frac{n}{5}+1)---\left(5\right)$

In formula, n is collecting well group well number; C is iterations.

As shown in Figure 3, step is as follows successively for genetic algorithm for solving flow process: the objective matrix that given drilling well sequential optimization is corresponding; The mathematical model that given drilling well sequential optimization is corresponding; Select genetic algorithm control parameter (population size N, iterations NG); The N number of initial individuals of stochastic generation; Build the fitness that fitness function calculates each individuality; Judge that genetic algorithm is the need of termination, if iterations >=NG, then directly export preferred result, terminate; If iterations < is NG, then continue to perform selection opertor (Propertional model), perform crossover operator (partial mapped crossover method) and perform mutation operator (exchange variation method), then repeat to judge that genetic algorithm is the need of termination step.

As shown in Figure 4, step is as follows successively for crossover operator process flow diagram: given two parent individualities (given by previous step selection opertor); Given intersection bandwidth; Generate two original filial generations; Determine two halved belt mapping relations; Generate two offspring individuals; Calculate the fitness of offspring individual; Crosscheck (if filial generation is not as parent, then gives up filial generation and get parent).

(10) the multiple optimum solutions (the anti-collision well section total length of whole well group is minimum) provided for genetic algorithm and suboptimal solution, in conjunction with batch drilling well requirement, optimize the drilling well order being conducive to anti-collision and batch drilling well most.

In above-mentioned steps (2), well interpolation method can by existing document 1 (Han Zhiyong. inclination drilling Design calcu-lation (second edition) [M]. Dongying: publishing house of China University Of Petroleum Beijing, 2007.) the well interpolation method that provides.

In above-mentioned steps (4), anti-collision scan analysis can by existing document 1 (Han Zhiyong. inclination drilling Design calcu-lation (second edition) [M]. Dongying: publishing house of China University Of Petroleum Beijing, 2007.) the anti-collision scan method that provides.

In above-mentioned steps (5), well track error analysis can by existing document 2 (Liu Gonghui, this capital of Dong, Gao Deli. error ellipsoid (circle) and well are handed over and are touched probability analysis [J]. drilling technique, 2000,23 (3): 5-12.) the well track error analysis method provided in.

In above-mentioned steps (6), venture analysis is touched in well friendship can by existing document 3 (B.Poedjono, G.Akinniranye, G.Conran, etal.Minimizingtheriskofwellcollisionsinlandandoffshored rilling [R] .SPE108279,2007) the well separation coefficient computing method provided in.

The present invention proposes using whole collecting well group anti-collision well section total length minimum as drilling well sequential optimization index first, by establishing target matrix, build the step such as drilling well sequential optimization model and genetic algorithm for solving, draw multiple optimum solution (the anti-collision well section total length of whole well group is minimum) and suboptimal solution, combine batch drilling well again to require (usually adopting the clinodiagonal drilling well order of jumping " day " word or sphere of movements for the elephants), thus select the drilling well order being conducive to anti-collision and batch drilling well most.The present invention roughly step is as follows: after the Hole clean scheme of given whole well group, carries out interpolation calculation (spacing 1m) successively, coordinate transformation, minimum distance scanning analysis, well track error analysis (error of calculation ellipsoidal parameter), well hand over and touch venture analysis (calculate well separation coefficient), set up drilling well order selected objective target matrix, genetic algorithm for solving, comprehensively preferably (take into account batch drilling well requirement) to whole designed path.

Be explained below in conjunction with certain marine collecting well group drilling well sequential optimization example.

For the marine collecting well group of THE WESTERN SOUTH CHINA SEA.This well group has 8 mouthfuls of orientation wells, and design orientation and whipstock point (KOP) are in table 1, and well head groove Rankine-Hugoniot relations is shown in Fig. 5 (spacing 2.28m).

Table 1

Pound sign	Whipstock point	Design orientation
			A1 well	650m	72.11°
A2 well	550m	79.20°
			A3 well	2250m	69.57°
A4 well	1000m	201.57°
			A5 well	550m	258.33°
A6 well	750m	329.84°
			A7 well	650m	33.27°
A8 well	500m	70.23°
			Pound sign	Whipstock point	Design orientation
A1 well	650m	72.11°
			A2 well	550m	79.20°
A3 well	2250m	69.57°
			A4 well	1000m	201.57°
A5 well	550m	258.33°
			A6 well	750m	329.84°
A7 well	650m	33.27°
			A8 well	500m	70.23°

By Track desigh scheme given in drilling engineering design book, get anti-collision scan separation delta L=1m, genetic algorithm for solving is adopted (to get initial population N=32 by the marine cluster wells drilling well order optimization of mathematical models of the method establishment that provides above, halved belt width w=3, iterations C=100), result of calculation is as follows.

$\begin{array}{cc}{P_1}^{*}=\&lsqb;7,4,1,5,2,6,8,3\&rsqb;,& 1031m\end{array}$

$P_2^{*}=\&lsqb;7,4,1,5,6,2,8,3\&rsqb;,1031m$

$P_3^{*}=\&lsqb;7,4,1,5,6,8,2,3\&rsqb;,1031m$

$P_4^{*}=\&lsqb;7,4,1,5,6,3,8,2\&rsqb;,1031m$

$P_5^{*}=\&lsqb;7,4,1,5,8,6,2,3\&rsqb;,1039m$

$P_6^{*}=\&lsqb;7,4,1,5,8,2,6,3\&rsqb;,1039m$

$P_7^{*}=\&lsqb;7,4,1,5,2,8,6,3\&rsqb;,1039m$

$P_8^{*}=\&lsqb;7,4,5,1,6,2,8,3\&rsqb;,1056m$

$P_9^{*}=\&lsqb;7,4,5,1,6,8,2,3\&rsqb;,1056m$

Above-mentioned optimum solution and anti-collision well section total length corresponding to suboptimal solution closely, can ignore its difference, then combine batch drilling well and require preferred drilling well order.According to the groove of well head shown in Fig. 5 Rankine-Hugoniot relations, known drilling well order and meet batch drilling well requirement completely, all the other several drilling well orders also basic satisfied batch drilling well require (only having 1 needs rig to be moved to adjacent well head groove).Consider anti-brushing up against and criticize drilling well requirement, the first-selected drilling well order of this well group , be secondly drilling well order , be drilling well order again with

In addition, also can calculate this well group anti-collision well section total length maximal value is 1462m, and than the recommendation drilling well many 431m of order (41.80%), drilling well order corresponding with it has at least following several:

P ₁＝[3,1,7,2,8,6,5,4]，1462m

P ₂＝[3,1,7,8,6,2,4,5]，1462m

P ₃＝[3,1,2,4,7,8,6,5]，1462m

P ₄＝[1,2,3,7,8,4,6,5]，1462m

P ₅＝[1,2,3,7,8,6,4,5]，1462m

Can find out, above-mentioned several drilling well order is all difficult to meet anti-collision and batch drilling well requirement.Such as, A3 well location is in well group center and whipstock point the darkest (2250m), this well ought to be finally bored by cluster wells anti-collision construction principle, if this well of subdrilling certainly will cause the anti-collision well section total length of whole well group larger, be also difficult to meet batch drilling well requirement (having 2 needs at least by drilling-machine shifter extremely adjacent well head groove).

In addition, the drilling well order that existing drilling engineering design provides is [8,5,2,7,1,6,4,3], is also A8 → A5 → A2 → A7 → A1 → A6 → A4 → A3.By the present invention give method, can obtain with it corresponding anti-collision well section total length is 1156m, than having recommended drilling well order many herein 125m (12.12%).Also find, this drilling well order meets batch drilling well not yet completely and requires (having 1 needs by drilling-machine shifter extremely adjacent well head groove) simultaneously.

Above-mentioned instance analysis shows, the marine cluster wells drilling well sequence optimizing method that the present invention provides advantageously, obviously can reduce marine collecting well group anti-collision well segment length, reduces anti-collision difficulty of construction and workload, also contributes to batch drilling mode.

Claims (3)

1. a marine collecting well group drilling well sequence optimizing method, is characterized in that comprising the following steps:

(1) the Hole clean scheme of given whole well group;

(2) adopt well interpolation method, interpolation calculation is carried out to whole designed path, and replaces original designed path with the track after interpolation processing;

(3) whole designed path is transformed into in certain reference point coordinate system that is initial point;

(4) according to anti-collision scan method, minimum distance scanning analysis is carried out to whole designed path;

(5) whole designed path is composed with measuring error, according to well track error analysis method, calculate the error ellipsoid parameter that whole interpolation point is corresponding;

(6) according to well separation coefficient computing method, the whole interpolation point of this well designed path well separation coefficient corresponding with all the other each wells is calculated by mouth well;

(7) all interpolation point numberings well separation coefficient on each designed path being less than set-point are stored in S set respectively _ijin, i, j=1,2 ..., n, finally arranges out objective matrix S, sees formula (1);

$S=\left[\begin{array}{cccccc}\mathrm{\&phi;}& S_{12}& ...& S_{1j}& ...& S_{1n}\\ S_{21}& \mathrm{\&phi;}& ...& S_{2j}& ...& S_{2n}\\ \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}\\ \\ \end{array}& \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}\\ \\ \end{array}& \begin{array}{c}.\\ .\\ .\end{array}\\ S_{i1}& S_{i2}& ...& S_{ij}& ...& S_{in}\\ \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}\\ \\ \end{array}& \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}\\ \\ \end{array}& \begin{array}{c}.\\ .\\ .\end{array}\\ S_{n1}& S_{n2}& ...& S_{nj}& ...& \mathrm{\&phi;}\end{array}\right],(i,j=1,2,...,n)---\left(1\right)$

In objective matrix S, each element is a subclass, S set _ijbe with i well for stratigraphic well, during with j well for comparing well, on i well designed path, separation coefficient is less than whole node serial numbers of set-point; The S set as i=j _ij=φ, φ are empty set;

(8) be numbered from 1 to n n mouth well, replace pound sign by numbering, each drilling well order all correspondence one has the fully intermeshing of n element; If fully intermeshing P is any one drilling well order, i.e. P=[p ₁, p ₂..., p _k... p _n] (k=1,2 ..., n; 1≤p _k≤ n); Set up drilling well sequential optimization model so that the anti-collision well section total length of whole well group is minimum for optimization aim, see formula (2);

$L=min\underset{k=2}{\overset{n}{\&Sigma;}}card\left(\underset{m=1}{\overset{k-1}{\&cup;}}{S}_{p_k{p}_m}\right)\&CenterDot;\mathrm{\&Delta;}L---\left(2\right)$

In formula, L is the anti-collision well section total length of the whole well group that drilling well order P is corresponding, m; represent k-1 set union, identical element not double counting; Card (*) represents the number of element in set; Δ L is borehole track interpolation point spacing, m;

(9) genetic algorithm for solving drilling well sequential optimization model is adopted;

(10) the multiple optimum solution provided for genetic algorithm and suboptimal solution, in conjunction with batch drilling well requirement, optimize the drilling well order being conducive to anti-collision and batch drilling well most.

2. one according to claim 1 marine collecting well group drilling well sequence optimizing method, it is characterized in that, in step (5): first calculate the error ellipse of perpendicular hole section correspondence and the circle of uncertainty of equivalence one-tenth area equation, then calculate error ellipsoid corresponding to lower curved section.

3. one according to claim 1 marine collecting well group drilling well sequence optimizing method, it is characterized in that, in step (9): described genetic algorithm comprises to be selected initial population and iterations, structure fitness function and optimizes genetic operator, and genetic operator comprises selection opertor, crossover operator and mutation operator; Specific as follows:

(1) initial population: N group sequence is produced at random for n mouth well individual, as initial population, get initial population N=4n;

(2) fitness function: by formula (3) structure fitness function

$f\left(i\right)=1-\frac{L\left(i\right)-L_{min}}{L_{max}-L_{\min }+\mathrm{\&Delta;}L}---\left(3\right)$

In formula, f (i) is i-th individual corresponding adaptive value in initial population; L (i) is i-th individual corresponding anti-collision well section total length, m; L _maxand L _minbe respectively maximal value and the minimum value of anti-collision well section total length in initial population, m; Δ L is borehole track interpolation point spacing, m;

(3) selection opertor: initial population is selected in adoption rate system of selection;

(4) crossover operator: each drilling well order all correspondence one has the fully intermeshing of n element, adopts partial mapped crossover method, is avoid parent to intersect to produce poor filial generation, also adds crosscheck step; Crossover operator basic step is as follows:

1. according to well number n, the width w of halved belt is determined by formula (4), then the reference position of random selecting halved belt;

$w=\mathrm{int}\left(\frac{n}{5}\right)+2---\left(4\right)$

2. exchange the halved belt of parents, form original filial generation;

3. the mapping relations of two halved belts are determined;

4. according to mapping relations, the non-crossing band of original filial generation is changed, until filial generation does not conflict;

(5) mutation operator: adopt and exchange variation method, to a certain drilling well order Stochastic choice two positions, then exchanges the pound sign on these two positions mutually; In order to avoid precocious phenomenon, carry out pressure populational variation equally, and add variation checking procedure prevent variation produce poor filial generation;

(6) iterations: determine iterations by formula (5);

$C\&GreaterEqual;50\&times;\mathrm{int}(\frac{n}{5}+1)---\left(5\right)$

In formula, n is collecting well group well number; C is iterations.