CN109763810A - A kind of reservoir gas drilling risk recognition methods - Google Patents
A kind of reservoir gas drilling risk recognition methods Download PDFInfo
- Publication number
- CN109763810A CN109763810A CN201811594758.7A CN201811594758A CN109763810A CN 109763810 A CN109763810 A CN 109763810A CN 201811594758 A CN201811594758 A CN 201811594758A CN 109763810 A CN109763810 A CN 109763810A
- Authority
- CN
- China
- Prior art keywords
- accident
- parameter
- drilling
- vector
- feature vector
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Withdrawn
Links
Landscapes
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention belongs to drilling safety technical fields, disclose a kind of reservoir gas drilling risk recognition methods, comprising: the fuzzy recognition method of Drilling Troubles: establish model；Determine membership function；Establish fuzzy relation matrix；Y is sought according to fuzzy relation；Application of the grey correlation in Drilling Troubles identification.Present invention application is fuzzy and grey correlation methods realize the identification of gas drilling complex accident；According to the complex working condition multi-parameter feature monitoring data of reading, the drilling failure recognition result alarm of maximum possible is provided；The loss of complex situations bring is minimized；For this engineering problem of the risk control of gas drilling reservoir, a possibility that its security risk occurs is reduced to the greatest extent in conjunction with present information acquisition, data processing, risk intelligent recognition and administrative skill, by Basic Theory Analysis, model foundation, simulation analysis of computer, the method and software realization of the gas drilling identification of accidental events of observation prediction in real time are established.
Description
Technical field
The invention belongs to drilling safety technical field more particularly to a kind of reservoir gas drilling risk recognition methods.
Background technique
Currently, the prior art commonly used in the trade is such that
Drilling well is a hidden underground engineering, there is a large amount of ambiguity, randomness and uncertain problem, due to
To unclear or subjective consciousness the incorrect decision of understanding of objective circumstances, many complex situations can be generated or even cause serious thing
Therefore less serious case expends a large amount of manpower and material resources and time, severe one leads to the discarded of full well.According to drilling data analysis in recent years, drilling well
In the process, the time for handling complex situations and drilling failure accounts for about the 6~8% of construction total time, and one possesses hundred bench drill machines
Oil field just has 6~8 bench drill machines doing idle work in 1 year, the consumption and time of getting along well let alone fund are proportional, but want
It is much bigger, this be how surprising waste.The prevention of challenge and drilling failure and the people for the treatment of method are not understood, inevitably not
Encounter these problems, and once having encountered can be panic, manner is undeliverable, minor illness is controlled into serious disease, serious disease controls into dead disease.And
Understand the prevention of challenge and drilling failure and the people for the treatment of method, know what's what if encountering these problems, takes just
True measure can often turn danger into safety, pull through, this is only the mature prerequisite condition of drilling well worker.
Air drilling nineteen fifty-three originates from the U.S., is to replace boring using gas, gas-liquid mixture fluid as circulatory mediator
The drilling technology of well liquid.Air drilling is started to walk evening at home, and Sichuan starts sporadicly to attempt in 50~seventies of last century, and 80
Age end Xinjiang office import China first set air drilling equipment, has started the field test of air drilling.Pass through more than ten years
Exploration, to there is very fast development after 2000, the rapid development of Sichuan Basin gas drilling, gas drilling after 2005
The characteristics of having shown revolutionary technological progress presents the great potential of the technology.But the country is generally still in test at present
The popularization stage.For difficult discovery, difficult exploitation, the difficult point that drilling well is difficult, drilling well is slow, positive development test underbalance (gas) drilling well skill
Art.Its development substantially experienced three phases: 50~seventies of 20th century, exploratory stage early stage；1999~2004, starting
Stage；2005 so far, developing stage.
The advantages of air drilling: the damage to payzone is greatly reduced, exploration and development benefit is improved；Leakage is eliminated to brill
The influence of well operations；It can get higher rate of penetration, reduce the totle drilling cost of drillng operation (Jing Yueshen, effect are more obvious)；
Drilling life of bit can be improved；Operating condition and environment become more to clean.
The limitation of gas drilling: being confined to the interval of wellbore stability, and caving ground easily causes bit freezing；It is big to tackle stratum
It is limited to measure water, fuel-displaced ability；Well control risk is bigger.
It cannot accomplish the drilling failure recognition result alarm of maximum possible in the prior art；It cannot be complex situations bring
Loss minimizes；The method and software realization of the gas drilling identification of accidental events of observation prediction in real time are not established.
In conclusion problem of the existing technology is:
It cannot accomplish the drilling failure recognition result alarm of maximum possible in the prior art；It cannot be complex situations bring
Loss minimizes；The method and software realization of the gas drilling identification of accidental events of observation prediction in real time are not established.
Summary of the invention
In view of the problems of the existing technology, the present invention provides a kind of reservoir gas drilling risk recognition methods, this hairs
It is bright to be achieved in that a kind of reservoir gas drilling risk recognition methods, the reservoir gas drilling risk recognition methods packet
It includes:
The fuzzy diagnosis of Drilling Troubles: million feature vector models are established；Determine membership function；Establish fuzzy relation matrix；
Accident pattern Y is sought according to fuzzy relation；
Drilling Troubles are identified using grey correlation: each parameter in Accident Characteristic vector is normalized；It asks
Detection coefficient and each accident criterion feature vector, X outiAfter the degree of association of (1≤i≤m), obtain relating sequence [R]=
[r1,r2,...,rm], when the degree of association of parameter to be checked and a certain accident criterion feature vector is more than to give in normal state in advance
Fixed λiWhen (1≤i≤m), which occurs corresponding class accident, realizes Classification and Identification.
Further, sign feature vector is
Accident pattern Y includes:
Y1: stratum produces gas；Y2: borehole well instability；Y3: stratum produces water；Y4: well shooting；
Y5: drilling string failure；Y6: stratum output pernicious gas；Y7: land equipment and personnel safety；
Failure Characteristic Parameter Xi(i=12 ... 12) it is respectively as follows:
x1- injection gas pressure；
x7- pipeline walls thickness value；
x2- hook load；
x3- torque；
x4- return out gas flow
x5- return out gas pressure
x6- return out gas relative humidity
x7- pipeline walls thickness value；
x8- ground is flammable, concentration of toxic gases；
x9- return out chip morphology variation；
x10- return out landwaste amount；
x11- acquisition image abnormity；
x12When-brill；
It determines in membership function, [0,1] is divided into many subspaces by fuzzy logic, respectively indicates xiIt is under the jurisdiction of event A's
In various degree, referred to as membership function muiA(x), it is denoted as μA:X→[0,1]；
Membership function is determined in drilling process, Dynamic Signal constantly changes, with characterization parameter increase or subtract
Small, positive number expression parameter increases, negative number representation parameter reduces, and 0 indicates no any variation, xiThe person in servitude of faults severity
Membership fuction are as follows:
a1: represent the upper limit of a certain parameter；a'1: represent the upper limit of a certain parameter；a'2: represent the lower limit of a certain parameter；
a2: represent the lower limit of a certain parameter.
Further, fuzzy relation matrix:
It is asked in Y according to fuzzy relation, fuzzy relation equation Y=XR brings sign vector X and fuzzy relation matrix into mould
Paste relation equation is calculated, wherein it is recognition result that component the maximum, which corresponds to accident pattern, in Y.
Further, include: to Drilling Troubles identification using grey correlation
The objective system accident pattern number of identification is m, and the element number that the feature vector of every kind of accident is included is n,
Establish corresponding accident criterion pattern character vector matrix:
Every a line in matrix represents the feature vector of a certain accident above, and each element in each column is having the same
Physical significance；
The feature vector of actually measured parameter are as follows:
YT=[y1,y2,...,yn]
The feature vector Y of actually measured parameterTWith standard feature vector matrix XRIn pass between each characteristic standard vector
Connection degree determines which kind of accident occurs；
Vector Y to be checkedTWith standard feature vector matrix XRIn between each Accident Characteristic standard vector corresponding element minimum absolutely
Error amount and worst error value are respectively as follows:
Vector Y to be checkedTWith standard feature vector matrix XRIn each characteristic parameter degree of association coefficient formula are as follows:
ρ in formula is resolution ratio, and 0 < ρ < 1 defines vector Y to be checked after ρ value takes calmlyTWith Accident Characteristic vector XiPass
Connection degree are as follows:
Different characteristic parameters is different to the influence degree of system mode, and different weights is assigned to each characteristic parameter
ωj, the degree of association are as follows:
The weights sum of each characteristic parameter is 1, that is, is met:
Before calculating correlation, each parameter in Accident Characteristic vector is normalized；Find out detection coefficient
With each accident criterion feature vector, XiAfter the degree of association of (1≤i≤m), one group of relating sequence [R]=[r can be obtained1,
r2,...,rm], when the degree of association of parameter to be checked and a certain accident criterion feature vector is more than given in normal state in advance
λiWhen (1≤i≤m), which occurs corresponding class accident, realizes Classification and Identification.
Another object of the present invention is to provide a kind of computers for realizing the reservoir gas drilling risk recognition methods
Program.
Another object of the present invention is to provide a kind of Information Numbers for realizing the reservoir gas drilling risk recognition methods
According to processing terminal.
Another object of the present invention is to provide a kind of computer readable storage mediums, including instruction, when it is in computer
When upper operation, so that computer executes the reservoir gas drilling risk recognition methods.
In conclusion advantages of the present invention and good effect are as follows:
A kind of reservoir gas drilling risk recognition methods of present invention offer provides one kind and is based on instrument detection data,
The identification of gas drilling complex accident is realized using fuzzy and grey correlation methods；It is supervised according to the complex working condition multi-parameter feature of reading
Measured data provides the drilling failure recognition result alarm of maximum possible；Identification example shows to carry out risk with grey relational grade pre-
The method of survey is feasible, and has the characteristics that calculation amount is small, simple and convenient, result is reliable；
A kind of reservoir gas drilling risk recognition methods of present invention offer, by the Parameters variation for analyzing drilling failure
The modern fault identification technology of properties study and mathematics approach application to drilling failure identify in method propose with feasible
Property method this for improve Drilling Troubles identification accuracy be it is very necessary, complex situations bring loss be down to most
Lower bound degree.
A kind of reservoir gas drilling risk recognition methods of present invention offer, for the risk control of gas drilling reservoir
This engineering problem contracts to the greatest extent in conjunction with present information acquisition, data processing, risk intelligent recognition and administrative skill
A possibility that its small security risk occurs is established and is seen in real time by Basic Theory Analysis, model foundation, simulation analysis of computer
Survey the method and software realization of prediction gas drilling identification of accidental events.
Detailed description of the invention
Fig. 1 is underground fish schematic diagram provided in an embodiment of the present invention.
Fig. 2 is the preceding underground fish schematic diagram of sidetracking provided in an embodiment of the present invention.
Fig. 3 is the deep 1 new wellbore of well sidetracking of duck provided in an embodiment of the present invention and former borehole position schematic diagram.
Fig. 4 is provided in an embodiment of the present invention 73/ 4 " with brill top jar plane of disruption schematic diagram.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to
Limit the present invention.
Reservoir gas drilling risk recognition methods provided in an embodiment of the present invention are as follows:
The fuzzy diagnosis of Drilling Troubles: million feature vector models are established；Determine membership function；Establish fuzzy relation matrix；
Accident pattern Y is sought according to fuzzy relation；
Drilling Troubles are identified using grey correlation: each parameter in Accident Characteristic vector is normalized；It asks
Detection coefficient and each accident criterion feature vector, X outiAfter the degree of association of (1≤i≤m), obtain relating sequence [R]=
[r1,r2,...,rm], when the degree of association of parameter to be checked and a certain accident criterion feature vector is more than to give in normal state in advance
Fixed λiWhen (1≤i≤m), which occurs corresponding class accident, realizes Classification and Identification.
The fuzzy recognition method of Drilling Troubles provided by the invention are as follows:
In drilling process, overwhelming majority situations have certain signs before drilling failure occurs, and sign is exactly certain parameters
Abnormal change.A certain sign may cause the possibility that a variety of accidents occur, and a kind of generation of accident may be drawn by a variety of signs
It rises, can solve the problem of such complexity with fuzzy mathematics diagnosis,
Assuming that:
1) generation of certain accident not will lead to the generation of other accidents.
2) sign that may be showed when drilling well may have 16 kinds, be expressed as x1,x2,...,x11,x12.12 kinds of sign difference
It is:
3) failure being likely to occur has 9 kinds, is expressed as y1,y2,...,y6,y7.7 kinds of accidents are respectively: stratum produces gas, well
Wall unstability, stratum produce water, well shooting, drilling string failure, stratum output pernicious gas, land equipment and personnel safety.
The brief step of fuzzy theory:
1) sign feature vector is constructed It is that object has sign xiDegree of membership.
2) fuzzy relationship matrix r is established.
3) according to fuzzy relation Y=XR, Y, fault feature vector are solved It is pair
As with failure yjDegree of membership.
Fuzzy theory specific method:
1) sign feature vector is constructed
It is contemplated that following common Drilling Troubles:
Y1: stratum produces gas Y2: borehole well instability Y3: stratum produces water Y4: well shooting
Y5: drilling string failure Y6: stratum output pernicious gas Y7: land equipment and personnel safety.
According to the above various failure sign analyses, following Failure Characteristic Parameter is selected:
Xi(i=12 ... 12) it respectively indicates:
x1- injection gas pressure x7- pipeline walls thickness value
x2- hook load x8- ground is flammable, concentration of toxic gases
x3- torque x9- return out chip morphology variation
x4- return out gas flow x10- return out landwaste amount
x5- return out gas pressure x11- acquisition image abnormity
x6- return out gas relative humidity x12When-brill
Table one: engineering parameter involved by analysis ratiocination rule occurs for every safety failure
Illustrate: ★ is adequate condition, can qualitative determination；◆ for auxiliary decision condition, possible response lag or it is not responding to
2) membership function is determined
0,1 two-valued function is generalized to the continuous logic of arbitrary value in desirable [0,1] closed interval by fuzzy mathematics, with corresponding
Degree of membership a possibility that failure occurs is described, [0,1] is divided into many subspaces by fuzzy logic, they respectively indicate xiIt is subordinate to
Belong to the different degrees of of event A, referred to as membership function muiA(x), it is denoted as μA:X→[0,1]。
There are many kinds of the methods for determining membership function, and based in drilling process, Dynamic Signal constantly changes, with it
Characterization parameter increases or reduces, and indicates that failure develops towards serious development degree, the increase of positive number expression parameter, negative number representation
Parameter reduces, and 0 indicates no any variation, xiThe membership function of faults severity may be defined as:
a1: represent the upper limit of a certain parameter；a'1: represent the upper limit of a certain parameter；a'2: represent the lower limit of a certain parameter；
a2: represent the lower limit of a certain parameter.
Threshold value will be set according to the concrete condition of well, for subsequent calculating, be given below various parameters threshold value and
Actually detected value, see the table below:
| The upper limit | The upper limit | Normal value | Lower limit | Lower lower limit | |
| x1 | 100 | 90 | 87 | 85 | 80 |
| x2 | 18.1 | 17.5 | 17.3 | 17.2 | 16.85 |
| x3 | 2100 | 2050 | 2000 | 1950 | 1900 |
| x4 | 17.5 | 17.0 | 16.6 | 16.4 | 15.7 |
| x5 | 1.98 | 1.93 | 1.90 | 1.86 | 1.80 |
| x6 | 3400 | 2780 | 2720 | 2600 | 2570 |
| x7 | 0.98 | 0.87 | 0.85 | 0.80 | 0.70 |
| x8 | 250 | 200 | 180 | 150 | 100 |
| x9 | 250 | 200 | 180 | 150 | 100 |
| x10 | 0.098 | 0.093 | 0.090 | 0.088 | 0.083 |
| x11 | 1.1 | 1.0 | 0.99 | 0.96 | 0.91 |
| x12 | 33 | 30 | 28 | 27 | 18 |
Due to upper table data when provide at random, so by the data processing of upper table, i.e. normalized, with each ginseng
Numerical value is gone divided by respective maximum value, and data are as follows after normalization:
| The upper limit | The upper limit | Normal value | Lower limit | Lower lower limit | |
| x1 | 1.0000 | 0.9000 | 0.8700 | 0.8500 | 0.8000 |
| x2 | 1.0000 | 0.9669 | 0.9558 | 0.9503 | 0.9309 |
| x3 | 1.0000 | 0.9762 | 0.9524 | 0.9286 | 0.9048 |
| x4 | 1.0000 | 0.9714 | 0.9486 | 0.9371 | 0.8971 |
| x5 | 1.0000 | 0.9747 | 0.9596 | 0.9394 | 0.9091 |
| x6 | 1.0000 | 0.8176 | 0.8000 | 0.7647 | 0.7559 |
| x7 | 1.0000 | 0.8878 | 0.8673 | 0.8163 | 0.7143 |
| x8 | 1.0000 | 0.8000 | 0.7200 | 0.6000 | 0.4000 |
| x9 | 1.0000 | 0.8000 | 0.7200 | 0.6000 | 0.4000 |
| x10 | 1.0000 | 0.9490 | 0.9184 | 0.8980 | 0.8469 |
| x11 | 1.0000 | 0.9091 | 0.9000 | 0.8727 | 0.8273 |
| x12 | 1.0000 | 0.9091 | 0.8485 | 0.8182 | 0.5455 |
For sign feature vector can be calculated, the detected value of one group of hypothesis is given below.
(85,17.4,1800,16.5,1.95,2720,0.75,190,260,0.085,1.3,28)；
Then according to membership function formula, license feature vector is calculated
X=(, 0, -1,0,0.4,0,0.5,0,1,0,4,1,0)
3) fuzzy relation matrix is established
The foundation of fuzzy relation matrix is the key that Drilling Troubles identification, follows actual conditions in order to objective, considers different
Characteristic parameter determines each thing generally according to a large amount of data, the experience of expert and statistical analysis to the ability of identification of accidental events
Therefore the degree of membership of corresponding each reason is allowed to more to react close between sign and failure to construct fuzzy relationship matrix r
Degree.According to table one to construct fuzzy relation matrix.
4) Y is asked according to fuzzy relation
Since fuzzy relation equation Y=XR brings (1) formula sign vector X and (2) formula fuzzy relation matrix into fuzzy pass
It is that equation is calculated, wherein accident pattern corresponding to component the maximum is recognition result in Y.
Application of the grey correlation provided by the invention in Drilling Troubles identification are as follows:
Using gray relative analysis method, a kind of drilling failure recognition methods based on grey relational grade is proposed.
If the objective system accident pattern number to be identified is m, the element number that the feature vector of every kind of accident is included
For n, then corresponding accident criterion mode (or reference model) eigenvectors matrix can be established:
Every a line in matrix represents the feature vector of a certain accident above, and each element in each column is with identical
Physical significance, but the size of its value may be different.
If the feature vector of actually measured parameter are as follows:
YT=[y1,y2,...,yn] (3)
Therefore, it is necessary to research be actually measured parameter feature vector YTWith standard feature vector matrix XRIn each spy
The correlation degree between standard vector is levied, so that it is determined which kind of accident has occurred.
Define vector Y to be checkedTWith standard feature vector matrix XRIn corresponding element is most between each Accident Characteristic standard vector
Small absolute error value and worst error value are respectively as follows:
Vector Y to be checkedTWith standard feature vector matrix XRIn each characteristic parameter degree of association coefficient formula are as follows:
ρ in the formula is resolution ratio, it can preset, and value range is 0 < ρ < 1, but by ρ value general
In the case of take into 0.5 be it is less reasonable because incidence coefficient is influenced by resolution ratio ρ, different ρ values is not it sometimes appear that
The same degree of association, as ρ=0.5, even if the degree of association can also be greater than 0.333 ..., therefore will basis without any association between two vectors
Practical problem carries out correct value to ρ.
After ρ value takes calmly, vector Y to be checked can be definedTWith Accident Characteristic vector XiThe degree of association are as follows:
Since influence degree of the different characteristic parameters to system mode is different, in order to show the important of a certain characteristic parameter
Property, it needs to assign each characteristic parameter different weight ωj, then the degree of association just can be changed to:
The weights sum of each characteristic parameter is 1, that is, is met:
In actual application, since the measurement unit of different parameters is different, there is the differences in dimension and the order of magnitude
It is different, therefore before calculating correlation, it needs that each parameter in Accident Characteristic vector is normalized；Find out detection system
Several and each accident criterion feature vector, XiAfter the degree of association of (1≤i≤m), can be obtained one group of relating sequence [R]=
[r1,r2,...,rm], when the degree of association of parameter to be checked and a certain accident criterion feature vector is more than to give in normal state in advance
Fixed λiWhen (1≤i≤m), so that it may think that such accident has occurred in the system, to reach the Classification and Identification to mode.
Consider following common drilling failure:
NX1- stratum produces gas
NX2- borehole well instability
NX3- stratum produces water
NX4- down-hole blasting
NX5- drilling string failure
NX6- stratum output toxic and harmful gas
NX7The failure of-land equipment
According to the analysis to above-mentioned 7 kinds of drilling failure features, we select following Accident Characteristic parameter to describe accident spy
Sign:
x1- sediment outflow pipeline returns out gas (O2) component and concentration Change in Mean
x2- sediment outflow pipeline returns out gas flow and pressure Change in Mean
x3- hook load and bit pressure Change in Mean
x4- injection gas pressure Change in Mean
x5The opposite variation of-drill bit zero dimension torque mean value
x6- sediment outflow pipeline returns out gas relative humidity Change in Mean
x7- sediment outflow pipeline returns out landwaste amount and granularity changes
x8- above propose drilling tool overpull value
x9- decentralization drilling tool is hampered value
x10- pipeline wall thickness reduction
x11- acquisition image abnormity
x12The toxic gas concentration variation in-well site
Each accident pattern and corresponding Parameters variation table
Since the parameter of foregoing description Accident Characteristic is difficult to be indicated with the size of a certain specific value, in actual conditions
In, the opposite situation of change of numerical value is concerned, therefore a threshold value can be set, when the increase and decrease amount of parameter is more than threshold value
When, that is, think that parameter is changed.Simply increased with 1 expression parameter, -1 expression parameter reduces, and is become with 0.5 expression parameter
Change fluctuation, is not changed with 0 expression parameter.
In each standard failure mode feature vector, since the importance of each characteristic parameter is different, it can assign not
Same weight coefficient.When two or more accident pattern has similar features, join in certain distinguishing features
Certain weight ω should be assigned on number.
According to analysis before, the value of ρ needs to judge according to concrete condition, needs to follow following principle:
(1) ρ should be according to the concrete condition dynamic value of actual observation vector；
(2) when singular value occurs in observation vector, ρ should take lesser value, to overcome the dominating role of singular value；
(3) when observation vector is more steady, ρ should take biggish value, fully demonstrate the globality of the degree of association.
And in parameter attribute vector Y to be checkedTWith standard feature vector matrix XRIn the variation of each characteristic standard vector comparatively
Than more gentle, therefore the value of ρ can be greater than 0.5, be 0.8 by its value.
According to above-mentioned analysis, it can establish following mode standard Accident Characteristic vector, see the table below:
Mode standard Accident Characteristic vector machine corresponds to weight
Illustrate: being used in tableThe characteristic parameter of expression is unrelated with corresponding accident pattern or can ignore its shadow
Loud characteristic parameter is replaced in actually calculating with 0；In order to protrude the main feature of accident, ignore secondary feature, in table
Weight setting to certain relevant parameters is 0.
Determine the feature vector of parameter to be checked, it is necessary to determine first each in the case that zero defects occurs in drilling process
The a reference value of characteristic parameter generates the identification of mistake in order to eliminate system because of the influence of accidentalia, can continuously use 3~5
A reference value of the secondary recognition result as final recognition result, then at the parameter real-time measurement in some time interval
Reason extracts the value of a characteristic parameter in parameter vector to be checked.It is each in standard parameter feature vector, parameter attribute vector sum to be checked
After the weight coefficient of characteristic parameter determines, the degree of association between parameter vector to be checked and each standard parameter feature vector can be calculated
R (i), when r (i) is greater than or equal to critical degree of association rε(i) when, then corresponding accident has occurred.All kinds of drilling failures it is critical
The degree of association see the table below:
The critical degree of association of all types of accidents
| Number | Accident pattern | The critical degree of association |
| NX1 | Stratum produces gas | 0.6 |
| NX2 | Borehole well instability | 0.55 |
| NX3 | Stratum produces water | 0.7 |
| NX4 | Down-hole blasting | 0.67 |
| NX5 | Drilling string failure | 0.71 |
| NX6 | Toxic and pernicious gas output | 0.64 |
| NX7 | Land equipment and personnel safety | 0.58 |
Below with reference to embodiment, the present invention is further described.
Embodiment 1, the deep 1 well damper of Qinghai Qaidam duck fracture accident
Deep 1 well of duck is that Chaidamu Basin, Qinghai Province the northern fringe fault block is set with the pile of stones, earth or grass beam-a bite of duck lake structural belt duck lake structurally
The prospect pit of well depth 4950.00m is counted, bores and meets Youshashan group, lower Youshashan group and upper dried firewood ditch stratum in Shizigou group, casing programme
For surface layer, two layers of protective casing and production tail pipe, ZJ-70D drilling machine operation is used.
1:00 on May 7 in 2001 creeps into first layer protective casing well using φ 444.5mmPDC drill bit, pendulum assembly structure
For section to when the layer of oil-sand mountainous region, weight on hook drops to 83T by 106T on 1706.15m, pump pressure drops to 9.6MPa by 10.9MPa,
Torque decline.Find that φ 197mm fractures with top jar top connection ontology stress-relief groove is bored after trip-out.Calculate top of fish depth
1592.38m fish length 113.77m.
As shown in Figure 1, fish in well are as follows: φ 444.5mmPDC drill bit × 0.54m+730 × 730 connectors × 0.51m+ φ
229mm drill collar × 26.16m+ φ 443mm stabilizer × 2.00m+ φ 229mm drill collar × connector of 25.93m+731 × 630 ×
0.38m+ φ 203mm drill collar × 52.73m+ φ 197mm is with brill bumper sub × 5.13m+ top jar ontology connector 0.39m.
Accident treatment in two stages, by 28 end of day in May totally 21 days, 2,100,000 yuan of economic loss.
(1) salvaging on 7 days~May 10 May first stage fish salvages situation and is shown in Table 1.In salvaging, top of fish is not being calculated
Encounter fish, determine to use φ 444.5mm rock bit fish finding top, as a result differs 4.67m, practical top of fish with calculating top of fish
1597.05m, judging to fracture with brill bumper sub falls in drill collar annular space.
As shown in Fig. 2, lower die nipple salvages fish by card, back-off is impregnated, as a result again bores die nipple safety joint, φ 127mm
Bar × 1 piece, drilling rod × 3 piece φ 139.7mm, meter 40.27m fall in well, fish overall length 154.04m in well, and top of fish is deep
1556.78m determines barefoot laterally drilling.
(2) second stage 11~May of May 28, electrical measurement, cementing plug, sidetracking (sidetracking situation is shown in Table 2).To determine sidetracking
Well section has carried out electrical measurement hole deviation and orientation, at 1.826 ° of hole deviation of 1450.00m~1480.00m well section maximum, 156 ° of orientation~
145.8 °, determine that Orientation differences point 1480.00m is skew back point.Cement plug well section 1400.00m~1554.00m is designed thus, it is real
Border cement surface is 1313.35m, and after extra cement plug is bored, using+2 ° of bent subs of screw rod, 255 ° of orientation is (with former wellbore axis side
90 ° or so of position) directional sidetracking.
The safety that well section construction in lower part must be taken into consideration when selection side tracking point orientation and bent sub degree, if side tracking point side
Position is consistent with former borehole axis orientation, and the new wellbore formed after sidetracking is easy to bore back former wellbore, to avoid such case, it is necessary to select
With the bent sub compared with lordotic, then biggish hole deviation may be generated in kickoff point (KOP) well section.If side tracking point orientation and former borehole axis
Orientation is opposite (180 °), and the new wellbore formed after sidetracking is not easy to bore back former wellbore.Also the bent sub of smaller angle may be selected, but
It will form severe dog-leg in side tracking point well section, influence the drillng operation of lower part well section.Therefore, the orientation of side tracking point selects Ying Yuyuan
Wellbore axis oriented perpendicular.
The practical sidetracking time since May 22, to totally 7 day time May 28.
As shown in figure 3, drilling depth 213.00m is total to from sidetracking well depth 1480.00m sidetracking to 1693.00m, and 1.0 ° of hole deviation, orientation
45 °, with former wellbore wheelbase from 4.1m, sidetracking success restores normal drilling (wellbore and former borehole position are shown in Fig. 3 after sidetracking).
As shown in figure 4, φ 197mm has 7mm deep with the half for boring jar lower contact stress-relief groove fracture perimeter of section
Shiny surface, (crackle) is suspected of defect caused by heat treatment.Look into product number, top jar 86049, bumper sub 86003,
It for 86 years products, is detected before sending well by pipe station, outbound is sent well (detection project is unknown) after qualification, and situ visualization is not
Used product.It is identified with boring jar 1984,1986 are initial production, even if being not used, also store 14 years as long as,
Inner seal aging already.Therefore, the accident belongs to downhole tool quality accident.
Come due to failing to salvage with brill bumper sub, it is unclear that bumper sub fracture state moves down 4.67m points according to top of fish
Analysis, fracture location also should be at brill bumper sub lower contact stress-relief grooves.Fracture process is first fractureed with brill bumper sub mandrel,
Moment is inserted into lower part top of fish (φ 203mm drill collar) annular space with bumper sub main body is bored after fractureing, and will sprain with brill top jar lower contact disconnected,
It sprains to have no progeny and falls into pit shaft annular space with brill bumper sub.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (7)
1. a kind of reservoir gas drilling risk recognition methods, which is characterized in that the reservoir gas drilling risk recognition methods
Include:
The fuzzy diagnosis of Drilling Troubles: million feature vector models are established；Determine membership function；Establish fuzzy relation matrix；According to
Fuzzy relation seeks accident pattern Y；
Drilling Troubles are identified using grey correlation: each parameter in Accident Characteristic vector is normalized；Find out inspection
Survey coefficient and each accident criterion feature vector, XiAfter the degree of association of (1≤i≤m), relating sequence [R]=[r is obtained1,
r2,...,rm], when the degree of association of parameter to be checked and a certain accident criterion feature vector is more than given in normal state in advance
λiWhen (1≤i≤m), which occurs corresponding class accident, realizes Classification and Identification.
2. reservoir gas drilling risk recognition methods as described in claim 1, which is characterized in that sign feature vector is
Accident pattern Y includes:
Y1: stratum produces gas；Y2: borehole well instability；Y3: stratum produces water；Y4: well shooting；
Y5: drilling string failure；Y6: stratum output pernicious gas；Y7: land equipment and personnel safety；
Failure Characteristic Parameter Xi(i=12 ... 12) it is respectively as follows:
x1- injection gas pressure；
x7- pipeline walls thickness value；
x2- hook load；
x3- torque；
x4- return out gas flow
x5- return out gas pressure
x6- return out gas relative humidity
x7- pipeline walls thickness value；
x8- ground is flammable, concentration of toxic gases；
x9- return out chip morphology variation；
x10- return out landwaste amount；
x11- acquisition image abnormity；
x12When-brill；
It determines in membership function, [0,1] is divided into many subspaces by fuzzy logic, respectively indicates xiIt is under the jurisdiction of the different journeys of event A
Degree, referred to as membership function muiA(x), it is denoted as μA:X→[0,1]；
Membership function is determined in drilling process, Dynamic Signal constantly changes, with increasing or reducing for characterization parameter, just
Number expression parameter increases, negative number representation parameter reduces, and 0 indicates no any variation, xiFaults severity is subordinate to letter
Number are as follows:
a1: represent the upper limit of a certain parameter；a′1: represent the upper limit of a certain parameter；a′2: represent the lower limit of a certain parameter；a2: generation
The lower limit of a certain parameter of table.
3. reservoir gas drilling risk recognition methods as described in claim 1, which is characterized in that fuzzy relation matrix:
It is asked in Y according to fuzzy relation, fuzzy relation equation Y=XR brings sign vector X and fuzzy relation matrix into fuzzy pass
It is that equation is calculated, wherein it is recognition result that component the maximum, which corresponds to accident pattern, in Y.
4. reservoir gas drilling risk recognition methods as described in claim 1, which is characterized in that using grey correlation to drilling well
Fault identification includes:
The objective system accident pattern number of identification is m, and the element number that the feature vector of every kind of accident is included is n, is established
Corresponding accident criterion pattern character vector matrix:
Every a line in matrix represents the feature vector of a certain accident above, and each element physics having the same in each column
Meaning；
The feature vector of actually measured parameter are as follows:
YT=[y1,y2,...,yn]
The feature vector Y of actually measured parameterTWith standard feature vector matrix XRIn association journey between each characteristic standard vector
Degree determines which kind of accident occurs；
Vector Y to be checkedTWith standard feature vector matrix XRIn between each Accident Characteristic standard vector corresponding element minimum absolutely accidentally
Difference and worst error value are respectively as follows:
Vector Y to be checkedTWith standard feature vector matrix XRIn each characteristic parameter degree of association coefficient formula are as follows:
ρ in formula is resolution ratio, and 0 < ρ < 1 defines vector Y to be checked after ρ value takes calmlyTWith Accident Characteristic vector XiThe degree of association
Are as follows:
Different characteristic parameters is different to the influence degree of system mode, and different weight ω is assigned to each characteristic parameterj, close
Connection degree are as follows:
The weights sum of each characteristic parameter is 1, that is, is met:
Before calculating correlation, each parameter in Accident Characteristic vector is normalized；Find out detection coefficient and every
One accident criterion feature vector, XiAfter the degree of association of (1≤i≤m), one group of relating sequence [R]=[r can be obtained1,
r2,...,rm], when the degree of association of parameter to be checked and a certain accident criterion feature vector is more than given in normal state in advance
λiWhen (1≤i≤m), which occurs corresponding class accident, realizes Classification and Identification.
5. a kind of computer program for realizing reservoir gas drilling risk recognition methods described in Claims 1 to 4 any one.
6. a kind of information data processing for realizing reservoir gas drilling risk recognition methods described in Claims 1 to 4 any one
Terminal.
7. a kind of computer readable storage medium, including instruction, when run on a computer, so that computer is executed as weighed
Benefit requires reservoir gas drilling risk recognition methods described in 1-4 any one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811594758.7A CN109763810A (en) | 2018-12-25 | 2018-12-25 | A kind of reservoir gas drilling risk recognition methods |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811594758.7A CN109763810A (en) | 2018-12-25 | 2018-12-25 | A kind of reservoir gas drilling risk recognition methods |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109763810A true CN109763810A (en) | 2019-05-17 |
Family
ID=66450270
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811594758.7A Withdrawn CN109763810A (en) | 2018-12-25 | 2018-12-25 | A kind of reservoir gas drilling risk recognition methods |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109763810A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113062731A (en) * | 2019-12-30 | 2021-07-02 | 中石化石油工程技术服务有限公司 | Intelligent identification method for complex working conditions under drilling well |
| CN113919669A (en) * | 2021-09-26 | 2022-01-11 | 德联易控科技(北京)有限公司 | Method and device for determining risk information of risk control object |
-
2018
- 2018-12-25 CN CN201811594758.7A patent/CN109763810A/en not_active Withdrawn
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113062731A (en) * | 2019-12-30 | 2021-07-02 | 中石化石油工程技术服务有限公司 | Intelligent identification method for complex working conditions under drilling well |
| CN113062731B (en) * | 2019-12-30 | 2024-05-21 | 中石化石油工程技术服务有限公司 | Intelligent identification method for complex underground drilling conditions |
| CN113919669A (en) * | 2021-09-26 | 2022-01-11 | 德联易控科技(北京)有限公司 | Method and device for determining risk information of risk control object |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200240243A1 (en) | Method for intelligently determining hydrate drilling and production risks based on fuzzy judgment | |
| EP2917127B1 (en) | Well integrity management using coupled engineering analysis | |
| CN105089620B (en) | Monitoring system, the method and device of bit freezing | |
| BR112015009197B1 (en) | METHOD AND SYSTEM FOR PERFORMING A DRILLING OPERATION | |
| US11091989B1 (en) | Real-time parameter adjustment in wellbore drilling operations | |
| CN107292467A (en) | A kind of drilling risk Forecasting Methodology | |
| CN107292754A (en) | A kind of drilling risk forecasting system | |
| CN109763810A (en) | A kind of reservoir gas drilling risk recognition methods | |
| Krishna et al. | Conventional and intelligent models for detection and prediction of fluid loss events during drilling operations: A comprehensive review | |
| CN111734399A (en) | Intelligent leakage blocking method and system in drilling process | |
| Naraghi et al. | Prediction of drilling pipe sticking by active learning method (ALM) | |
| CN115586086A (en) | Borehole wall instability analysis method based on big data | |
| CN115841247A (en) | Digital drilling risk monitoring method and device | |
| US11655701B2 (en) | Autonomous torque and drag monitoring | |
| Zhao et al. | An adaptive pattern recognition method for early diagnosis of drillstring washout based on dynamic hydraulic model | |
| CN104632207A (en) | Method for measuring collapse influence factors of hole wall of cast-in-situ bored pile | |
| Lakirouhani et al. | Hydraulic fracturing breakdown pressure and prediction of maximum horizontal in situ stress | |
| Wang et al. | Dynamic wellbore stability analysis based on thermo-poro-elastic model and quantitative risk assessment method | |
| CN113550741A (en) | Method for detecting minimum inner diameter of casing | |
| CN107448187B (en) | Underground measuring device | |
| CN114036703B (en) | Method for analyzing strength of deep well petroleum casing pipe of salt-gypsum rock stratum | |
| JPH07301574A (en) | Three-dimensional crustal stress analyzing method | |
| Yang et al. | Fault Diagnosis Method and Application of ESP Well Based on SPC Rules and Real‐Time Data Fusion | |
| Zhao et al. | Early detection of drillstring washout based on hydraulics model and pattern recognition method | |
| Wen et al. | Intelligent prediction and key factor analysis to lost circulation from drilling data based on machine learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190517 |