Full field reservoir modeling of shale assets using advanced data-driven analytics

Soodabeh Esmailia, Shahab D. Mohagheghb, *



aAsset  Development Team, North Operation, California Resources Corporation, California 90024, USA

bWest Virginia University, 345-E Mineral Resources Bldg., P. O. Box 6070, Morgantown, WV 26506, USA



Hydrocarbon production from shale has attracted much attention in  the recent years. When applied to this prolific and hydrocarbon rich resource plays, our understanding  of  the complexities of  the flow mechanism (sorption process and flow behavior in complex fracture systems-induced or natural) leaves much to be desired. In this paper, we present and discuss a novel approach to modeling, history matching of hydrocarbon production from a Marcellus shale asset in southwestern  Pennsylvania using advanced data mining, pattern recognition and machine learning technologies. In  this new approach instead of imposing our understanding of the flow mechanism, the impact of multi-stage hydraulic fractures, and the production process on the reservoir model, we allow the production history, well log, completion and hydraulic fracturing data to guide our model and determine its  behavior. The  uniqueness of  this technology is  that it  incorporates the so-called “hard data” directly into the reservoir model, so  that the model can  be  used to optimize the hydraulic fracture process. The  “hard data” refers to field measurements during the hydraulic fracturing process such as  fluid and proppant type and amount, injection pressure and rate as  well as  proppant concentration. This  novel approach contrasts with the current industry focus on  the use of  “soft data” (non-measured, interpretive  data such as  frac  length, width, height and conductivity) in  the reservoir models. The  study focuses on  a Marcellus shale asset that includes 135 wells with multiple pads, different landing targets, well length and reservoir properties. The full field history matching process was successfully completed using this data driven approach thus capturing the production behavior with acceptable accuracy for individual wells and for the entire asset.



1. Introduction



Much of  the success in  turning the shale source rock  into an economically viable and producible hydrocarbon reservoir is accredited to George Mitchell and his  team of geologists and engineers at  Mitchell Energy & Development1.  The  success in  production of shale oil and shale gas dates back  to 1981 when multiple combinations  of   processes  and  technologies  where  examined before ultimately succeeding in 1997 with the use of a “slick-water” frac   that  made production from Barnett Shale economical and changed the future of  the US natural gas  industry (NGW,  2011). Today  horizontal wells that include multi-stage, multi-cluster hydraulic fractures and pad drilling are  the norm in developing shale oil   and  shale  gas    assets  in   North  America  and   expanding throughout the world.


Shale reservoirs  are   characterized by  extremely low   permeability rocks that have a  number of  unique attributes, including high organic content, high clay  content, extremely fine  grain size, plate-like micro-porosity, little to  no  macro-porosity, and coupled Darcy  and Fickian flow through the rock  matrix. Unlike conventional and even tight sandstone gas reservoirs where all the gas is in the free   state in  the pore space, the gas  in  shale is  stored  by compression (as free  gas)  and by adsorption on the surfaces of the solid  material, either organic matter or minerals (Guo  et al., 2012).

This combination of traits has  led  to the evolution of hydraulic fracture stimulation involving high rates, low-viscosities, and large volumes of  proppant. The  stimulation design for  plays such as Marcellus Shale is drastically different than anything else  that has been performed in  the past. It takes large amounts of space, materials, and equipment to  treat the Marcellus Shale to its  fullest potential (Houston et al.,  2009). Currently,  the Marcellus shale, covering a  large area in  the northeastern US, is  one of  the most sought-after shale-gas resource play  in  the  United States. It has presumably the largest shale-gas deposit in the world, having a potentially prospective area of  44,000 square  miles, containing about 500  TCF of recoverable gas  (Engelder, 2009).

Figure 1.  Data available in  the dataset that include location and trajectory, reservoir characteristics, completion, hydraulic fracturing and production details.

This  geological formation was known for  decades to  contain significant amounts  of  natural  gas but  was  never  considered economical.  Uneconomic  resources,  however,  are  often transformed into marketable assets by technological progress (Considine et al., 2009). Advances in  horizontal drilling and multi-stage hydraulic fracturing have made the Marcellus shale reservoir a focal point for  many operators. Nevertheless, our  understanding of the complexities associated with the flow mechanism in  the natural fracture and its coupling with the matrix and the induced fractures, impact of  geomechanical properties and optimum design of  hydraulic fractures is still  a work in progress.

A vibrant and fast-growing literature that covers operational and technological challenges of production from shale oil and shale gas  is currently thriving. The  research includes all aspects of drilling, completion, and production as well as difficulties in formation evaluation/characterization, in  modeling macro- and micro-scales of fluid transport, and in developing reliable reservoir simulators. Understanding reservoir properties like lithology, porosity, organic carbon, water saturation and mechanical properties of  the rock, which includes stresses, and planning completions based on  that knowledge is  the key  to  production optimization. Therefore, the final objective is to increase our  ability to integrate laboratory and petrophysical measurements  with geochemical, geological, petrologic, and geomechanical knowledge, to  develop a more solid  understanding of shale plays and to provide better assessments, better predictions, and better models.

Figure 2. Marcellus shale AI-based Full-field history matching process.

Reservoir  simulation  has   played  an   important  role towards achieving the above mentioned stated goals (Mohaghegh, 2013a). However there  are   still   many  challenges  to   overcome  before reaching the stated goals.  Firstly,  the physics of fluid flow in shale rocks haven’t been fully  understood, and are  undergoing continuous development as  the industry learns more (Lee  and  Sidle, 2010).  Secondly,  full   reservoir  simulation  is  resource  intensive and   time   consuming.   Thirdly,  challenges encountered  when applying conventional reservoir simulation to  shale resources (Mohaghegh, 2013b)  could  be   solved with  pattern  recognition technologies (Mohaghegh, 2000a,b,c).

In this paper, we developed an Artificial Intelligence-based model that is conditioned to all available field measurements (e.g. production history,  measured  reservoir characterizations including geomechanical  and  geochemical properties)  as   well   as   measured hydraulic fracturing variables like  slurry volume, proppant amount and sizes,  injection rate etc. Such model has  the potential to provide operators with an  alternative to history-match, predict and assess reserves in  oil  and gas   producing  shale  reservoirs.  The  pattern recognition approach not only  has  a much faster turnaround time compared to numerical simulation techniques, but also  offers reasonable accuracy while incorporating all available data compared to analytical and numerical techniques that are  very  selective in the type of field measurements that they use. The integrated framework presented in this paper enables reservoir engineers to compare and contrast   multiple   scenarios  and propose field development strategies.

Figure 3. Three well types from a single pad.

2.  Top-Down modeling - pattern recognition based reservoir modeling



Artificial intelligence and data mining refers to  a collection of tools and techniques that provide the means for  finding patterns among non-linear and interdependent parameters involved in the shale  oil   and  shale  gas   development  process.  Interest  in   the research of pattern recognition applications has  spawned in recent years.  Popular  areas  include:  data  mining  (identification  of  a ’pattern’, i.e.,  a correlation, or  an  outlier in  millions of  multidimensional patterns), document  classification (efficient search  of text documents), financial forecasting, and biometrics.


Top-Down modeling, a recently developed data-driven reservoir modeling  technology  (ISI,  2014),  is defined as a formalized, comprehensive, multi-variant,  full-field, and  empirical reservoir model, which takes into account all  aspects of  production from shale including reservoir characterization, completion and hydraulic fracturing parameters as well  as production characteristics. Despite the common practice in  shale modeling using a conventional approach, which is usually done at the well level  (Strickland et  al.,  2011),  this  technique  is  capable  of  performing  history matching for all individual wells in addition to  full field by taking into account the effect of offset wells.

Figure 4. Log-log plot of production rate as  a function of time for  one of 135 wells in  the asset studied for  this paper.

There are  major steps in the development of a Top-Down shale reservoir model that is enumerated as follows:


  • Spatio-temporal  database  development; the  first step  in developing a  data  driven shale reservoir model is  preparing  a representative  spatio-temporal database  (data  acquisition and preprocessing). The extent at which this spatio-temporal database actually represents the fluid flow behavior of the reservoir that is being modeled, determines  the  potential  degree of  success in developing a successful model. The  nature and class  of the AI-based shale reservoir model is determined  by   the  source of  this  database. The   term  spatio-temporal defines the  essence of  this database and  is  inspired from the  physics that  controls this phenomenon  (Mohaghegh, 2011). An  extensive data mining and analysis process should be conducted at this step to fully understand the data that is housed in this database. The  data compilation, curation, quality control and preprocessing is one of the most important and time consuming steps in developing an  AI-based reservoir model.
  • Simultaneous training and history matching of the reservoir model; in  conventional numerical reservoir simulation the  base model will be modified to match production history, while AI-based reservoir modeling starts with the static model and tries to honor it and not modify it during the history matching process. Instead, we will analyze and quantify the uncertainties associated with this static model at a later stage in the development. The model development and history matching in  AI-based shale  reservoir  model  are  performed simultaneously during the training process. The  main objective is to make  sure  that  the  AI-based shale reservoir  model learns fluid flow behavior in the shale reservoir being modeled. The spatio-temporal database developed in the previous step is the main source of information for building and history matching the AI-based reservoir model.

Figure 5. Inside and closest outside distance.

  • Simultaneous training and history matching of the reservoir model; in  conventional numerical reservoir simulation the  base model will be modified to match production history, while AI-based reservoir modeling starts with the static model and tries to honor it and not modify it during the history matching process. Instead, we will analyze and quantify the uncertainties associated with this static model at a later stage in the development. The model development and history matching in  AI-based shale  reservoir  model  are  performed simultaneously during the training process. The  main objective is to make  sure  that  the  AI-based shale reservoir  model learns fluid flow behavior in the shale reservoir being modeled. The spatio-temporal database developed in the previous step is the main source of information for building and history matching the AI-based reservoir model. In this work, an ensemble of multilayer neural networks is used (Haykin, 1999). These neural networks  are  appropriate for pattern recognition purposes in case  of dealing with non-linear cases. The neural network consists of one hidden layer with different number of  hidden  neurons,  which have been  optimized  based on   the number of  data records and the number of  inputs in  training, calibration and verification process (Mohaghegh, 2000a). It is extremely important to have a clear and robust strategy for validating the predictive capability of the AI-based reservoir model. The model must be validated using completely blind data that has not been used, in any  shape or form, during the development. Both training and calibration datasets that are  used during the initial training and history matching of  the model are  considered non- blind. As noted by  Mohaghegh (2011),  some may argue that the calibration - also   known as  testing dataset -is  also   blind.  This argument has  some merits but if used during the development of the AI-based shale reservoir model can  compromise validity and predictability of  the model and therefore such practices are  not recommended.

Figure 6.  History matching result for  entire field by  using the maximum possible combination of parameters.

  • Sensitivity analysis  and  quantification  of uncertainties; during the model development and history matching that  was defined  above, the  static model is  not modified. Lack  of  such modifications may present a weakness of this technology, knowing the fact  that the static model includes inherent uncertainties. To address this,  the AI-based reservoir modeling workflow includes a comprehensive set  of sensitivity and uncertainty analyses. During this step, the developed and history matched model is thoroughly examined against a wide range of changes in reservoir characteristics and/or operational constraints. The  changes in pressure or production rate at each well are  examined against potential modification of any  and all the parameters that have been involved in the modeling process. These sensitivity and uncertainty analyses include single- and combinatorial-parameter sensitivity analyses, quantification of uncertainties using Monte Carlo  simulation methods and finally development of type curves. All these analyses can be performed on individual wells, groups of wells or for the entire asset.
  • Deployment of the model in  predictive mode; similar to any other reservoir simulation model, the trained, history matched and validated AI-based shale reservoir model is deployed in predictive mode in order to be  used for  reservoir management and decision making purposes.

Figure 7.  History matching result for  entire field in  optimum history matched model.

3.  AI-based reservoir modeling in Marcellus shale



The study presented in  this manuscript  focuses on  part  of Marcellus shale that  includes 135  horizontal wells within more than 40 pads. These horizontal wells have different landing targets, well lengths and reservoir properties. During the development of the Top-Down shale model all  available data including static, dynamic, completion, hydraulic fracturing, operational constraint, etc. have been used for  training, calibration, and  validation of  the model. A complete list of inputs that are  included in main data set for development of the base model is shown in Fig. 1.

The   data set   includes more  than  1200 hydraulic fracturing stages (approximately 3700 clusters of hydraulic fracturing). Some wells have up to 17 stages of hydraulic fracturing while others have been fractured with as  few  as  four  stages. The  perforated lateral length ranges from 1400 to 5600 ft. The total injected proppant in these wells ranges from a minimum of about 97,000 lbs  up  to  a maximum of about 8,500,000 lbs and total slurry volume of about 40,000 bbls  to 181,000 bbls.

Figure 8. Data-driven full  field reservoir model training, calibration, and verification cross plots.

The wells are completed in both upper and lower Marcellus. The porosity of upper Marcellus varies from 5  to 10% while its  gross thickness is measured to be between 43 and 114 ft. The total organic carbon content (TOC) of the upper Marcellus in this area is between 0.8 and 1.7%. The reservoir characteristics  of  lower  Marcellus include porosity in the range of 8e14%, gross thickness between 60 and 120ft, and TOC of 2-6%.



4.  Results and discussion



During the  training  and  history  matching  process using AI- based modeling approach inclusion and exclusion of multiple parameters were examined in  order to determine their impact on model behavior. Fig. 2 includes a flowchart that shows the evolution process of developing the AI-based Marcellus shale full  field reservoir model. It starts from the base model (where most of the parameters are  included as our  first shot) to  converge to  the best history matched model where  optimum  number  of  inputs are identified.

Figure 9. List of the inputs in  optimum history matched model.

4.1.  Impact of different input parameters



Base  Model- As illustrated in Fig. 2, the base model was built by  incorporating all  available data that are listed in Fig. 1. This model consists of all field measurements  including well locations, trajectories,  static  data,  completion, hydraulic  fracturing  data, production rates, and operational constraints.


Effect of Offset Well- In order to consider the effect of offset wells and taking into account for  any  well  interference, all afore-mentioned properties for  closest offset well were included in  the modeling.


Effect of  Different Well Types- Since  drilling multiple wells from a pad is a common practice in the most shale assets and given the fact  that the horizontal wells drilled from a  pad experience different interaction with their offsets, three types of laterals have been defined. Fig. 3 shows the configuration of the three types of wells drilled from a  single pad. Based on this definition a  new parameter was added to the dataset marked as  “Well-Type.” This parameter was assigned values such as 1, 2, or 3 in order to incorporate the “Well-Type” information.

Figure 10. The  best (top) and worst (bottom) history matched wells in  the optimum history matched model.

Following is a brief  description of the three “Well-Types” used in the data-driven full field model.


  • Type one  Lateral: This type of lateral has  no  neighboring laterals and does not share drainage area. It does not experience any  “Frac Hits2”  from wells in  the same pad (it  might experience Frac-Hit from a lateral from an offset pad) and its reach will  be as far as its hydraulic fractures.
  • Type  two  Lateral: The  type two lateral has  only  one neigh-boring lateral and therefore, it shares part of the drainage area and “Frac Hits”  are  possible from laterals in the same pad (it might experience Frac-Hit from a lateral from an  offset pad).
  • Type three Lateral: The  type three lateral is bounded by two neighboring laterals thus, the drainage area will  be  shared and “Frac Hits” are possible from both sides in the same pad.  If a type three lateral experiences a Frac-Hit from an offset pad, it will  be from a different depth.



Effect of Different Flow Regimes- There are usually two distinct flow regimes that can be observed in all the wells. The first flow regime corresponds to the production of the initial free gas in the fracture/pore spaces, which is immediately available for production and it may last a few days to a few months (Flow regimetype one). Most of the wells have been observed to exhibit transient linear behavior as the main flow regime (Flow  regime type two).

Figure 11. History matching error histogram for  optimum history matched model.

This  transient  linear behavior is  characterized by  a  one-half slope on  a  log-log plot of  rate against time. The  transient linear flow regime is expected to be caused by transient drainage of low-permeability  matrix  blocks into  adjoining fractures  (Bello and Wattenbarger,  2010). These two flow regimes where introduced in  the data-driven full  field model as  dynamic property.  Fig.  4 shows the two flow regimes for  one of  the wells from the asset described in this case  study.


Effect of Distances Between Laterals- In order to consider the impact of the location (distance from other laterals in the same pad and closest lateral from an  offset pad), two distances were defined and used in the data-driven full field model: (1) distance between laterals of the same pad, (2) distance to closest lateral of a different pad. This concept is illustrated in Fig. 5.

Figure 12.  Examples of history matching results-excellent matches.

As  mentioned earlier,  the  base model was developed with maximum number of inputs that were available. The results of the history matching of the base model are  shown in Fig. 6. This figure shows the history matching results for  the entire field.  It must be noted that in the data-driven full field reservoir model the history match is performed on a well  by well basis. In order to generate the history match for  the entire asset, as shown in Fig. 6, productions from individual wells are  summed (the field measurements as well as the Top-Down model) and are  plotted.

The top plot in this figure is the history match (monthly production, the left  y axis  and the cumulative production, the right y axis). The  bottom bar chart shows the well  count (number of producing wells). In the top plot,  the orange dots represent the actual monthly rate (normalized to protect confidentiality of data) for the entire field while the green solid  line  shows  the  AI-based model  results. The orange area represents the actual cumulative production (normalized  to protect confidentiality of data) while the green area corresponds to cumulative production generated by the AI-based model.



4.2.  Optimum history matched model



Although, the  history matching results  drived by  using the maximum combination of parameters (Fig. 6) have reasonable accuracy, it is preferable to  reduce the number of input variables in data-driven  models. As  the number of  inputs in  a  data-driven model increases, a certain level  control over the model behavior may be compromised. Accordingly, a history-matching process was derived with minimum combination of parameters that can/should be  used to  achieve an  acceptable history match results for  individual wells and for  the entire field (the total number of  inputs decreased from 103  to 38).

Figure 13. Examples of history matching results-good matches.

The  inputs that were removed and the reason for  removing these are  as follows:


  • The upper and lower Marcellus rock  properties are  drastically different so  averaging them destroys a  major geologic factor that affects both the fracing and the resulting the well  performance. The  lack  of  allocated production should not pose a problem and in  fact  the resulting AI model could be  used to estimate the allocation of each zone and checking that result against the know reservoir properties. In most areas, you  are supposed to see  lower production from the upper Marcellus. This  again could be  used as  an  additional validation of  the AI-based model.
  • The perforated lateral length and total stimulated lateral length were included in  the data set.  Since  the length of stimulated lateral is  always 100  ft  longer than the length of  perforated lateral,  therefore  the  total  perforated  lateral  length  was removed from the model.
  • Instead of including stage based hydraulic fracturing data, the total values for slurry volume, proppant amount, etc. were used in the optimized case. In addition, the average injection rate and pressure that were not changing considerably were removed.
  • Since  the inside and closest outside distances from an  offset were included for each individual well,  therefore there was no need to include these two distances for  offset well   consequently they were removed from the input data.


The  final history match result for  the optimized model was improved and showed acceptable match of monthly gas  rate and the cumulative production for the entire field (Fig. 7). For this case, 80% of the data was used for  training and 20% for  calibration and verification  (10%  for  each). Fig.  8  shows the  cross plot of  the training, calibration, and verification of  the data-driven full  field reservoir model, which shows R2 of 0.99, 0.97  and 0.97  for training, calibration and verification, respectively. In this figure, the x-axis is the predicated monthly gas rate by the model while the y-axis is the actual gas  production rate from the field.  Fig. 9 shows the list  of inputs that were used in optimum history matched model. Fig. 10 shows two wells with the best and the worst history matching results in the optimum history matched model. This figure shows that the erratic behavior displayed by the well in the graph on the right could not be  captured by the data-driven full field reservoir model, even though the trend was followed.

Figure 14.  Examples of history matching results-average matches.

4.3.  Error  Calculation



The error percentage of monthly gas  production rate for all the 135  wells was calculated using the following equation:

where, YitTDM is the predicted production by TDM (AI-based model);


Yi;t  is the actual field data;

ΔYim     is  the measured maximum change in  actual production data;

Nt(i) is the number of month of production.


Fig. 11  shows the histogram of error for  the optimum history matched model. In this model, 101  wells were matched with less than 10% error (excellent), 22 wells had errors between 10 and 20% (good), 6  wells had errors between 20  and 30% (average) and 6 wells had errors of  more than 40%  (poor).  Several example  of excellent, good, average and  poor history matching results  are illustrated in Figs. 12-15.



5.  Conclusion



In  this paper, development and results of  a  data-driven,  full field Marcellus shale reservoir model was discussed with the aim of  overcoming the current issues associated with numerical simulation and modeling of shale gas  reservoirs. The advantage of this technology is its capability of handling and incorporating hard data instead of a rigid representation of flow and transport mechanisms in shale reservoirs. When dealing with complex non-linear systems such as flow in shale reservoirs, the available hard data  could  identify  its   functional  relationship   using  pattern recognition.

The  full-field history matching was performed with acceptable  accuracy.  This  model can  be  used for  Marcellus shale wells and reservoir performance prediction and field development.

Figure 15.  Examples of history matching resultsepoor matches.




The authors would like  to thank RPSEA and U.S. Department  of Energy for partially funding this study. Authors would like to thank Range Resources Appalachia LLC for  providing the data. We  also thank  and  acknowledge Intelligent Solutions Inc. for providing IMagine and IDEA     software packages for  the development of the Top-Down models.






Bello, R.O., Wattenbarger, R.A., December 2010. Modeling and analysis of shale gas production with a skin effect. Journal of  Canadian Petroleum Technology  49 (12), 37-48.


Considine, T., Watson, R., Entler, R., Sparks, J., 2009. An  Emerging Giant: Prospects and Economic Impacts of  Developing the  Marcellus Shale Natural Gas   Play. State College: EME  Department, PSU.


Engelder,  T.,  2009. Marcellus 2008:  Report  Card on the  Breakout Year  for Gas Production in the Appalachian Basin. Fort worth Basin Oil and Gas.


Guo, J., Zhang, L., Wang, H.,  Feng, G., 2012.  Pressure transient analysis for  multi- stage fractured horizontal wells in shale gas reservoirs. Transport  in Porous Media 93,  635e653.


Houston, N., Blauch, M.,  Weaver,  D., Miller, D., O’Hara, D., 23e25 September 2009. Fracture-stimulation  in the Marcellus Shale-lessons Learned in Fluid Selection and Execution. SPE  125987,  SPE  Eastern  Regional Meeting, Charleston, West Virginia, USA.


Haykin, S., 1999. Neural networks: a comprehensive foundation, second ed.  Prentice Hall,  Upper Saddle Rever, New Jersey.


ISI, 2014. Intelligent Solutions, Inc.


Lee,  J., Sidle, R., 2010. Texas A&M “Gas-Reserves Estimation  in Resource Plays”. SPE Economics & Management Journal 2 (2).


Mohaghegh, 2000a. Virtual intelligence applications in petroleum engineering: part 1; artificial neural networks.  Journal of  Petroleum  Technology, Distinguished Author Series, September 2000, 64-73.


Mohaghegh, S.D.,  2000b. Virtual intelligence applications in petroleum  engineering: part 2; evolutionary computing. Journal of  Petroleum  Technology, Distinguished Author Series, October 2000, 40-46.


Mohaghegh, S.D., 2000c. Virtual intelligence applications in Petroleum engineering: part  3;  fuzzy logic. Journal of  Petroleum  Technology,  Distinguished  Author Series, November 2000, 82-87.


Mohaghegh, S.D.,  2011. Reservoir simulation  and modeling  based on artificial intelligence and data mining (AI&DM). Journal  of  Natural Gas  Science and Engineering 3,  697-705.


Mohaghegh, S.D., 2013a. Reservoir modeling of shale formations. Journal of Natural Gas  Science and Engineering 12, 22-33.


Mohaghegh, S.D., 2013b. A Critical View of  Current State of  Reservoir Modeling of Shale  Assets.  SPE  165713-SPE Eastern  Regional Conference  and  Exhibition. Pittsburgh, PA. 20-22 August 2013.


NGW, 2011. Genesis: the North Texas Barnett Shale Opens New Energy Era,  North American Shale Revolution. Natural Gas  Week, Special report Sept. e Dec.  2011.


Strickland, R., Purvis, D., Blasingame, T., 14e16 June, 2011. Practical Aspects of  Re- serves Determinations for  Shale Gas.  SPE-144357, North American Unconventional Gas  Conference and Exhibition, The  Woodlands, Texas.


Article Info


Article history:

Received 19 July  2014

Received in revised form

26 November 2014

Accepted 6 December 2014

Available online 23 January 2015


E-mail address:  (S.D.  Mohaghegh).

Peer-review under responsibility of  China University of  Geosciences (Beijing).


1Mitchell Energy & Development. Mitchell sold his company to Devon Energy in 2002 in a deal worth  $3.5 Billion.


2Frac Hit  is  a phenomena that is  encountered when producing shale assets. It happens when production from a given well is disrupted as a function of hydraulic fracturing activities in an offset well. In  a Frac  Hit  the water used during the hydraulic fracturing of  an offset shows up and sometimes even completely shuts-in the producing well.

1674-9871/   2015, China University of Geosciences (Beijing) and Peking University. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC- ND license (