Transcript for:
Reservoir Characterization Workshop Overview

oh my goodness assalamualaikum hi everyone welcome to our one day free workshop for reservoir characterization and modeling we will start by five minutes presentation about reservoir solutions then mr terry dean well start so for reservoir solutions it's a company specialized in providing technical studies and courses for oil and gas clients and professionals the company incorporates many professionals with diverse technical background related to oil and gas acceleration and production our services include technical reservoir studies field development planning reservoir static and dynamic modeling economic visibility studies technical support for academic researches like phd and master degrees also we are assisting in writing technical papers and publishing also we are providing technical petroleum courses by the industry experts also we provide mentorship programs our next course which will be in 20 30 23 on june reservoir characterization using poor analysis routine core analysis and special core analysis it will be one week hand is on workshop for the content of the course the first day will be on the cooling how to prepare a program how to decide the sidewalk or with conventional touring then the day two will be for the measurements uh analysis measurements like the fluid saturation porosity bulk volume and permeability then we will talk about the rock typing methods like the fz win land and so on then in day three we will go to special group analysis we will talk about the special core analysis measurements like which ability the electrical properties and the capillary pressure then we will talk about the saturation height modeling and how can we use the saturation fighting into the static model and we will we will also talk about the relative permeability how to normalize and the average relative permeability and so on then in five and day five we will talk about the the data or analysis data processing for for the reservoir modeling how to represent the rock types into 3d model how to populate the rook types into 3d models we will have a practical very very 100 percent practical course that will initiate from zero to hero you will initiate from the raw data for the current analysis until populating the properties from the core analysis to 3d static mode for our webinar today the instructor for our webinar today is mr tej dean mr terry dean pinissa is a result is oriented and accomplished than 15 years of experience in identifying the location quantity and quality of hydrocarbon proving triac record of success in conducting uncertainty analysis maximizing asset value implementing technical assessment and advising on field development plan optimization he is expertise in directing offshore drilling operations acquiring reservoir data minimizing geological risks and uncertainty and dividing and executing 2 phi 2 positive 2 probability model demonstrated ability to update and maintain 7 icg and 1 ina geo models analyze and integrate geological data establish logic frameworks and build long lasting relationships with staff he is widely recognized as hand his own leader strategic problem solver neutrality communicator and skilled coach actually mr dean has approved experience he worked into petronas and also in libyan oil companies for the agenda for our today webinar or workshop we will have three lectures the film structure will be about the reservoir characterization and the modeling objectives then static data integration and modeling it will last about 45 minutes then we will have 10 minutes for questions then we will have five minutes for a break then we will go to the second lecture which will be about the reservoir characterization of the flow and the conceptual model what is it and the lecture will last about 45 minutes then we will have 10 minutes for question and 5 minutes as a break then we will go to the third lecture which will be about the static modeling workflow and static dynamic iteration process and the lecture will last about 55 minutes and we will have five ten minutes also for questions then five minutes to conclude all the information that we took through our uh workshop for the webinar rules when you have any question you can ask it into the zoom chat and our instructor will answer it into the suitable time mute your microphone and camera in case that you wanted to have a certificate for attendance with that workshop please fill in the form that will be sent into the zoom chat the certificate will be received through one week after the workshop for our upcoming courses and webinars you can join it through our webinar and youtube also contact us through mail or whatsapp okay that you wanted to register to the our upcoming course you can register it from the link is that will be sent into the zoom chat thank you for all attendees and mr taj dean you can start yes thank you very much i will share my screen now can everybody see my screen yeah yeah you can start that's good okay okay okay so today our topic is one of the most important topics in the uh not oil industry only but in every industry and in exploration or development for uh natural resources uh today we will talk about the uh the rcm which is the abbreviation for reservoir characterization and modeling you know when we go to the fields when we were students in geology or reservoir engineers or geophysicist geophysicists we used to look to the nature we used to look to the nature as we as the sky we see in this slide so we start we think think about how i can how how i can take this nature with all the complexity and the complicated processes inside and everything and use it for myself how i can change it into a digital numbers that i can handle to make predictions and obtain good results that will be the from which i would benefit in my industry so simply there are three steps that we uh that oil companies or the geoscience experts follow first to understand uh you know when you go to the uh outcrops or when you drill wheels so you obtain data from the rock okay so you start first to understand understand this uh data what they are telling you so you are you are in the face of gathering information before even start thinking about building um building a model or even characterized so you need to obtain informations to understand when we go to the outcrop when we go to outcrops we measure thickness and we take samples to uh examine and to uh take to the lab to measure the petrophysical properties and uh and everything then we go to the next step which is the characterize okay so i i i now i characterize my i'll identify what uh what what type of the rock that i have available in my hand is it limestone is it cold is it a sandstone uh um what kind of what kind of structure is it anticline is it the syncline is at the fault so i start uh creating synthetic data that enable me to understand or to characterize my response so that if so on this in this stage i can draw draw what i can see i see a i see as a fault structure then i can draw it i can sketch it with my hand so because because now i understand what this data tells me so that i can characterize so i use my free hand to characterize what i see then i will move now to the model because uh i i moved from the face of characterization where i could first make sketches or 1d 2d figures for my uh for the for my reservoir okay or for or for the for for deformation i could um and i could understand and make some measurements for uh some properties and numbers uh such as the numbers for for thickness and distribution and the extension of the rock units then now i get into the next stage which is the model now i want to make a 3d model i i move from 1d 2d data into a 3d data which will help me to exploit the resources and get and get the maximum revenue from exploitation so now uh let me introduce the presentation summary uh first we will uh go through the reservoir modeling definition you know uh uh stage what we do is that we are building a reservoir model so what is the reservoir is one model independent we will go through there is one modeling definition and objectives and get to know wires of war modelling as such an essential and daily basis project for oil companies because now uh reservoir reservoir model is used for different projects not not only for for volumes assessment and recovery node there are a lot of uh objectives to do reservoir uh modeling such as studying the the mechanical properties of the of the rock even in eur projects in coal and methane projects so it's uh the there is a variable use use applications for reservoir modeling so we will go through the objectives then uh we will go to the objectives of your company uh why we will talk about this because uh rivers reservoir there is a war model represent the base or the solid floor on which field development projects are achieved uh it is it represents essential key element through um ftp or field development uh plan workflow to reach desired or desired or the desired objectives or targets of the company then we will go to uh talk about the reservoir simulation and the difference between the when we say the uh resource the difference between reservoir reservoir modeling or the reservoir model and geological model then we will go to one of the very important parameters is the master for uh working in field development plan or when you are working as part of the rcm which is a teamwork the teamwork is very essential or the key element for for accomplishing rcm successful rcm study and fdp plans because you know mastermind is represented by teamwork so any successful project would not reach the top and succeed without teamwork so in this part we will we will recognize the technical positions required to accomplish reliability of our modeling and tasks that that everyone must accomplish to generate valuable inputs into the project then we will talk about as as we said in the second part the cellular model which is the origin or the root of any model either the dynamic model or the geological model uh so we will uh define it we will talk about the specificities you know uh each uh object has certain specifies to be to be to be valuable to be uh so that we can apply it for our advantages so geocellular models or the static models they have a certain specificities that need to be followed and applied so that you can call the geocellular model reliable to for rcm workflow or for accomplishing field development then we will uh after that we will understand the main differences between the geological and reservoir model as both serve different objectives and different domains but both serve serve or represent the part of the rcm workflow then we will as we said we will talk about the specific specificity of the reliable singular models so first what's uh reservoir model definition is the bar model is uh is a great uh is a grid of cells which allow okay allow us to manage very important to two important things first the key heterogeneity uh yeah you you know any is any reservoir in nature uh it's rarely necessarily to find uh homogeneous unless you have uh a good a nice or cleaned sand uh clean sand bodies like the uh rivers like uh this uh the braided the stream for the braided like the braided rivers but this is not always the case so if i just open the white board in nature usually if you have for example if you have a reservoir of a nice shaped channel like this one okay and you have a well you drilled several weld so you have a well here and well here okay so you expect that you are since you are drilling in the same sand body so expect that you have that you have almost a homogeneous uh reservoir homogeneous in terms of uh faces in terms of uh reservoir properties as well but unfortunately this is not always the case because if you have a reservoir composed of different of different stacking bodies of sand uh with different with different devotional faces ranging from braided to mandarin to uh point bars and also so you it's not always expected to have it homogeneous so the reservoir model uh comes to uh we we um we build reservoir models to highlight or to show us the heterogeneity or the key heterogeneities that affect what effect affect the flow units which from which the hydrocarbon is coming into the will bores uh also we the the the reservoir models represent the lithofacies and petrophysical properties distribution consistency as as as i mentioned here uh some uh if you have a homogeneous uh body of sand you expect to have the same average porosity for example okay but in case let's say let's say in carbonate in carbonate it's no way that you can find the heterogeneity is always the address of carbonate so when you talk about carbonate you are talking about heterogeneity so representing the models to represent heterogeneity in carbon reservoir it's it's truly uh a challenge it's truly a challenge and it's not easy to find that but you need to highlight the distribution and and and what the differences that this heterogeneity causes in terms of faces and in terms of metaphysical properties sometimes you you have a sand also but what you have you have a different body of sand so you have a you have a channel that goes like this and you have another one uh crossing like this you have a third one that goes like that okay so you have a different set of but i saw some of these in seismic and it was very beautiful image but uh dealing with this but dealing with the tragedy it was a nightmare okay so uh going back to the presentation so as i said uh we we do reservoir we do is one model for this uh to to to manage and represent these very important two elements which is the heterogeneity and consistency of the data so uh what the reservoir model is is for the main objective of reservoir model uh actually is not to predict okay there is a confusion about the notion of there is a model actually it's not uh to predict uh what we should find the reservoir uh what we should find in the reservoir it's not the uh it's the first concern it's the it's the main target in the exploration phase so when we are drilling a well so let's say that there is a company drilled somewhere somewhere uh okay they found a discovery they found from the seismic that there's a nice structure somewhere okay so they have their fee their field goes this way okay and we have a nice structure if you take cross-section let's say that we have a fall goes this way and uh and the the company drilled fed the first exploration well actually the uh the the exploration well enables us to find what's in uh what's inside the reservoir is it oil gas or or nothing so the main so what i want to say is that the main objective from the vermont is not to predict what we should find the reservoir or else it's to anticipate to anticipate the dynamic behavior the dynamic behavior of the reservoir we are going through long process in rcm okay workflow just to understand how this sand body or the rock body that we see here when we put wells how these wheels will perfo how the reservoir will behave and produce the oil uh or or the hydrocarbon from inside this is why we build the reservoir model so the water model is mainly used for um to simulate uh field evolution through time so uh so we use it for uh uh anticipating or predicting uh with production or will behavior as well uh fluid movements within the reservoir as certain at certain time and also pressure evolution with time how how the reservoir after drilling after drilling this this exploration will and we okay uh started developing our field by drilling a lot of development walls here and there and there and there okay the model should tell us before we drill this development will the how the pressure will differentiate or how it how it will change through time this is what actually the most important objective for the for any reservoir model so the jio um reservoir or reservoir reservoir model is one of the most um okay talk about the reservoir model i'm talking about geo model now now we are talking about the static model okay the static model represents uh one of the most important phases in integrated reservoir study workflow when you are talking about rcm as we said earlier we are say we said understand okay characterize and model the model it has two phases as we will see later the phase one is the static model or the geo model okay and the second one is the reservoir or the dynamic model so what the geo model represents represent one of the most important phases in integrated is velocity workflow why because it integrates multiple data from different disciplines which include reservoir geometry there's void characteristics physical properties faces everything okay it also takes into account dynamic information so as a as a geo modeler when you are building a geo uh a static model or geo model you're taking consideration some dynamic information such as the production data which is represented by the which is presented by plt also you will consider the uh the xpt okay that uh assist you to define the uh that assists the geoscientists to define the fluids contact uh also the core data the core analysis porosity and permeability all these things should be integrated into the model also it provides a key heterogeneity uh modeling after all this after all the data being gathered and integrated it will enable us to make or build a robust uh model that reflect the heterogeneity inside the inside our reservoir based on which we can understand how the fluid will behave in until certain certain point in the from now until certain time in the future so actually in this power the the person who's in who's needed in this part is actually the geologist or the geo model who's in charge of this uh uh part now let's come we talk we talked about the the objectives of reservoir modeling the objective actually the objective reservoir modeling is part of the objective or it's referred to the objective of the company what the company want from the uh for modeling there is war when companies make new discoveries and they and they had all the information and data that enabled good understanding of the reservoir it's now time time to make the reservoir profitable to make money okay as much as it can so uh the new discovery fields uh start to be subjected to development projects by drilling more wells to improve uh production hence actually increases recovery so uh through through feed life companies tend to uh acquire more data on regular and plan time to monitor his war behavior and and steer their okay direct or steer their reservoir uh toward the maximum uh profit target uh so uh uh so this is i actually the uh the objective of company so they build so they build the reservoir model and they keep updating the reservoir model based on each uh target okay so when we say that and let's say that in 2000 let's say in 2002 the uh company made this discovery well and in 2002 2003 they drilled okay appraisal well okay the appraisal well job was to because uh to check the boundary of the field the extension of the oil water contact and so on now they are happy now they make their first ftp okay they make their first ftp which is the field first development project okay and they set number of uh they apply the rcm workflow okay uh and uh based on this uh on the results they decide how many wells they want to drill and how and how and how much these wells will produce until from 2000 let's say from 2003 until let's say 2010. and after and and and after this time uh companies may go revisit the uh model and update the date and update the uh the model using the data that that has that has been obtained from 2003 or 2010 or which i mean here the new wills that the development was that uh drilled okay because from 2003 2010 there are several worlds where several wills was drilled this is a first world second development development world for development and so on okay so all these wells will be integrated into the next ftp plan project or the next static model to update and enhance our understanding for the reservoir flow behavior so so to obtain the optimal profitability of pro project there are key informations that we need to acquire okay uh acknowledge uh uh acquire and most [Music] which is the uh ref which uh summarized that okay require the uh knowledge of three important uh factors first one is the volume in place okay which is the hours type how much hydrocarbon are there under our feet we have a we have a field okay before we do uh anything how uh how much oil and this is usually is calculated or primarily calculated from the first three or two wells uh from the discovery or exploration wells are and the appraisal wells as well okay so it gives us a probabilistic it's not a fixed volume it's probabilistic volume but at least we know how much initially uh under our feet and also uh how much will be recovered we are talking here about the reserves uh usually companies um tend to make several scenarios based on the reserves on reservoir heterogeneity pressure data and also the data adequacy and reliability uh there is a long process indeed and there is in the reservoir and there is one modeling part uh associated with the recoverable reserves because you know uh there is always uncertainty as we are dealing more wells and obtain more data this uncertainty reduces but it doesn't end uncertainty always remains so always there's a probabilistic run to calculate the recoverable reserves in several scenarios and also how the wells are expected to behave okay how how how how the daily production will be before i drill this uh develop the development well okay i'm planning to uh drill four or five wells how much this how much this wealth will provide me from the total volume that from the total reserve that i would produce so the expected dual performance is another target or the company need to be aware of before paying money for any ftp project or to apply any rcm workflow okay so now let's come to the uh integrated field development plan what why do why they call it integrated field development plan because it integrates it's a multi-disciplinary integration this means that integrate data from different resources and also uh multiple multiple members or expert expertise are from different disciplines are involved in this project so it's a huge project okay uh objective of any development plan is simply to increase production okay so we are the companies set as we said in the uh as a as i mentioned earlier that companies apply ftp projects at uh every period of time to increase the production to exploit the to exploit the reservoir to the maximum benefit so any so any development project okay generally goes through essential essential um four four phases faces one which is the 3d geological model okay this is the geological model you you input the geological data in inside the model put your physical data and you come up with the volume assessment okay and the volume assessment can help the uh the ceo or the uh or the ceo of the company to make business decisions about how we can exploit this uh uh this reservoir or this field because if it's economical then there will be uh decisions but business decisions and if it's not economical i don't think that it would it will be a good chance to decide the business decision regarding any economical this is for example after that when you when once the geological model is done it will be transferred to reservoir part for reservoir or dynamic model to simulate flow or to build flow simulation model so based on the simulation model okay the the company will be will be able to forecast okay or to predict the production because based on the production uh prediction uh company can set the expenditure the expenditure for uh the development which includes drilling new wells or and making that acquisition so uh how much they will uh specify as expensive expenditure for surface facilities which includes the pipelines the crude and gas storage water treatment uh all the surface facilities depend on how and how many wells will be drilled and the distance between these wells and also are they drilling are the producers or they are injectors okay as we said integrated field development project it's a multi-disciplinary uh multi-disciplinary workflow you have several data from different resources as we can see here in this table uh actually it's a summary uh the the the the the the table the table that you see table that you see it uh it takes more than um two slides maybe or three two three more slides but i try to uh summarize it uh so that it can be simple to understand because uh this is only the geological part that i'm showing here so still there is a part for uh the the the reservoir engineering part geophysics part so uh as as as i said any ftp project need data from different disciplines uh these data go through variable processes okay soto provides solid interpretation for reservoir quality and also the behavior it's a teamwork so you have uh data such as seismic you need you need data such as seismic uh seismic can be original or for taken from analog fields you can also use the uh wheel logs and uh course uh what do you need this data for for uh first to construct the uh geological geological models uh to build structure grade stratigraphy grade sediment scientological uh and also you need you need this data as well to as we said in the at the beginning of the presentation is to characterize okay to build a concept we will talk about the concept in the next slides which is in the characterization phase so this is this synthetic data or the studies or results okay is carried out by uh experts from disciplines such as euphysicists the geophysicists and the geologists peter physicists and his for engineers are uh in charge of carrying of integrating this data not only uh obtaining the data and no it's integration and also uh we we have the will test the data fluid sampling lab experiments uh course such as the ccl which is the core analysis scale data which is the special analysis all of this and also uh taking fluid samples for pvt for uh doing with test interpretation all this data is uh uh [Music] integrated and interpreted by uh by reservoir engineers geophysicists geologists and petrophysicists because as i said it's uh it's integration project not only uh individual each one works separately no everybody should work together so uh so what's this you hear about reservoir simulation so what's the reservoir simulation anyway uh is there a simulation it's the uh you can see you can see that it represents the the the end okay the in the product or the end stage of geological model so once the uh reservoir once the static model has been built okay so i have my static model um okay so once i have the static model built with all the uh integrated studies inside so i have the faces okay i have the porosity first here i have the k which is the permeability i have fluid saturations inside it inside all the cells good saturation i have the thickness okay i have type of the structure as well is it fault is it faulted and so once once all these data are integrated into my uh static model uh the static model will be transferred to reservoir to start building the dynamic model let's say why do we need dynamic model or the reservoir model to run simulation we will come to this in more details later so dynamic model or reservoir model so what they would first they they input their dynamic data okay from production and pressure and so on then they run simulation and they run simulation to predict the field behavior and see how much these wealth can produce i hope that it's uh uh understood okay uh simulation is uh okay simulation is a numerical process okay for uh dynamic input to produce production and injection history of the field what's the uh what's the outcome of all this long process is history match curve is to imagine for each uh dynamic parameter inside the reservoir including the uh oil production the gas oil ratio the pressure and so on uh all this uh the the final target of the reservoir simulation is to provide a history match that can tell us whether this reservoir model is reliable or no and that depends on the match degree between the input data and the output data so as as we said for the dynamic models okay what we the main inputs because most most of the inputs are in the geological and the geological model but in the uh dynamic models okay we use we need the protection and protection data or injection rates if it's if there's injection wells we need the pressure because pressure changes over time so it's very impor very important input into the simulation and volumes of produced oil and water so now we come to uh when i'm we come as one of the most important uh part of the workflow of rcm which is the [Music] teamwork okay and working with your colleagues uh as we said that fdp and rcm workflow it's a long process and it's a multi-disciplinary so you need different expertise from different disciplines such as the log analyst look analyst is the um is the one who carry out quantitative and qualitative vlog interpretation okay you he is the one who identified the fluid contacts from logs from xpt highlight essential physical parameters uh for robust petrophysic physical models so it's very so it's the output of log analysis is very important input into the reservoir model there's also the laboratory physicist here we maybe we can call it the core specialist who is the uh not the core specialist who is charge of core acquisition now we are talking about the person who is doing or measuring petrophysical properties especially positive mobility and the special core analysis data in the core lab so he's the one who provide the lab let's call it lab-based metaphysical data through core and uh fluids analysis uh then we come to the geophysicist geophysicist say that is the uh he's the one who's providing the uh the boundaries of the reservoir you cannot uh model you cannot model the reservoir without knowing the boundaries of reservoir the top and the bottom so geophysics is the expert to provide seismic data including but not limited to seismic horizons and false and other attributes related to phases and fractures and all these data so the inputs is very important that represent also the beginning of any workflow as we will see a static model modeling workflow as we will see later in the next slides then another one whose who's the sedimentologist dermatologist is the one who identify sedimentary faces the macro and the micro it's very important to identify the faces because the faces is the carriers of the physical properties each faces but not always have the certain characteristics uh so they have a certain or uh special physical uh properties that can be identi that can be identified through facial distribution so uh so the sedimentologist can use the core description to identify the macrophages use photography to uh identify microporous microphases and micro and porosity as well permeability and also can you carry out correlation among wells to uh using court uh coordinator to see or to tell us how the faces are distributed throughout the field uh he also built a conceptual scientological model that showed the faces distribution which is this model can be important input into the static or reservoir model then we come to the uh geochemist okay he's geologists who provide information about the hydrocarbon fluids compositions and organic matters and finally we come to the reservoir engineer several engineers actually his job starts once the geological input has been uh built and interpreted and built into a 3d static model he or she will be in charge of uh sorry i mean maybe ask where is the geologist in this in all this part the geologist he or she would be in charge of building the building that is the static is of our model by integrating all the data as we said as we said earlier okay so uh uh this slide summarize the uh integration workflow for any rcm uh for any rcm or var characterization uh uh reservoir calculation modeling project so we have as you can see we have the geo model uh entry in in 3d we have uh the orange bubbles okay which indicates the data that should be input into the geological model we have the petrophysics from from the core data we have the symlogical model that we talked about uh earlier and that will be provided by the sedimentologists okay fractures will be provided also by the geophysicist or also by the sedimentologist if we can spot the these fractures from core we have the logo interpretation from the log analyst okay we also have the reservoir attributes that usually provided by the geophysicist reservoir attributes or let's say seismic attributes is associated with seismic data that can help us to uh that can help the geophysicist to detect any irregularities uh in the reservoir so we can get attributes for structure for fractures and also for uh phases and that also uh for sure depends on the quality of the um of the seismic uh we have also the uh as uh uh we have also the then after that we move to the reservoir model the reservoir model is the uh uh as we said if you compare the the the number or the amount of data input its reservoir model uh is not comparable to jo to a geo model in terms of uh data because as we said earlier what we need is the protection and pressure data and if there is and fluid characteristics for pvt into the reservoir model while most or let's say 70 percent of the data are input into the 3d static model okay uh uh cellular models are we here this uh we always hear um static model geological model uh dynamic model simulation models so all these uh the root of all these models is the cellular model so what's serial models it's a schematic description a schematic description of not reservoir any object so you are define it as the skeleton okay like like the skeleton of a building when you go to uh to any when you see when you look to the buildings i took some of these nice pictures for uh uh different buildings and the skeleton of these buildings it's not uh complete it's just a skeleton so uh these buildings later it's like a reservoir later will be uh filled with the cement will be filled with paintings and hardwares furnitures windows and everything so it's similar to reservoir so first we uh what we do we represent our reservoir as a cellular grid or as a serial model okay contain a lot of cells that will be later uh filled or in which there will be the data will be integrated and input so why do we need the cylinder cellular uh model uh uh uh we to understand the complexity of reality you know at the beginning of the session we said how we can convert or how we can transform nature how we can change the complex okay how we can change our complex things to a simple that can help us to predict and make good results for our uh profit so to understand the complexity of reality okay uh the relationship between uh inputs uh the heterogeneity and frameworks and also to quantify reality which which here we are talking about the volumes as well and reservoir properties because we are not uh because there are several volumes used uh by the ceiling we use uh from applying the cylinder from from using the cellular models such as the not only the hydrocarbon volumes even the rock volumes poor volumes and all these volumes are essential outputs from any from any reservoir model or geocylinder model so uh this is uh two slides uh i'm sure that you are familiar with the gulf x goldfix it's a north sea field we i i i'm showing here two uh pictures for uh one for the geological model okay let's call the geocellular model or as a vermont or dynamic uh model so uh we have here uh an example for the uh [Music] geoserial model for porosity okay distribution you can see that the reservoir before modeling has been divided and we will see later in the workflow of the modeling uh how we how we create these cells uh and these cells is filled with the input okay from from the porosity input and after that uh how how this and how this um [Music] there's one model there will be uh also the cells filled with the protection data and the pressure data but uh uh sorry i don't i don't have uh this is actually for uh i didn't i didn't put a picture for uh uh assimilation uh great this is for faces i i was supposed to define it here as faces so sorry it's what uh it's a mistake so this is a grid for porosity and this is a great for phases we will see later how we can create these cells or build a silver model at the beginning of the static modeling workflow so this is another example for uh cellular models so we have uh cells distributed in 3d okay x in the x x and y and z direction now we will uh maybe we can have a break now or we still have a time no problem if you wanted to have a break no problem okay okay okay but let's give let's give first 10 minutes to to answer questions okay okay no problem so uh attendance if if anyone has any question please uh put into the zoom chat uh participants that were asking about the certificate uh i will send the link for certification uh in after the second lecture java okay uh and the certificate for the participant participation and the attendance okay so it's a time for questions so if anyone that has any question please put it into the junction any question any questions guys okay okay so uh i think that we have some questions we have three questions and he doesn't chat okay playing what kind of type uh okay uh i i will answer this question in the second part when we uh when we reach the end of the static modeling workflow okay rando yeah because you because here you are concerned about the uncertainty so so uh i will answer this question at the end of the part two when we are concluding with the uh when we reach the uh the final stage of the uh static model this is where we where we will handle the uncertainty and and talk about the sensitivity for reservoir parameters okay uh actually the thickness the thickness okay for uh osman the uh the thickness of one cell actually it uh depends uh it's a it's agreement usually between the geologists and it's a subject of agreement between geologist and uh and reservoir engineer why uh in geology uh we tend to capture uh the nature as it is okay for uh example let me explain on the uh explain in the whiteboard uh in geology uh we would like the model to be fine grid okay what does mean find great we need we need a grid that capture we need a grid that capture okay let's say that we have uh uh i i have this well here okay sorry i have this well okay and i have uh let's say uh faces okay this is faces i don't want to go into much details but this is a shell layer okay and this is another faces okay so we have facials one here okay this is facial one okay and faces two this is shade and this is facial 3. okay what what we prefer as a geologist as a journalist when i build a static model i need a cell size that capture that capture all the faces that i see so this means that i want cell okay to capture faces one okay okay so because because i need a cell that represents these faces in the in reality then i want another cell okay that capture faces two and also another one okay that capture it will be smaller maybe it will be smaller than this uh um this uh two cells okay to capture the shell layer because it's very important shades represent uh flow barrier okay in simulation so it's very important to be captured and also another one to capture faces tree okay now we are talking in terms of geology but when it comes to simulation things is different reservoir engineers uh always prefer the coarse grid why and then and then they go and they and they got the point because when you go when it comes to uh to the flow okay the flow of hydrocarbon okay into the poor hole uh is not it's not uh it's more pressure and permeability okay it depends more on pressure and permeability than on the faces i'm not saying that face is not controlling no it controls but the the the the production and the and the primability of this let's say the uh field uh or larger scale primability or pressure permeability of buildup depends uh control the flow of this uh hydrocarbon so what they prefer is that they preferred of course a coarser grid which contain a larger okay a larger cell okay that can represent the flow of hydrocarbon into the reservoir the course of grit the problem of the course grid with the geology is that sometimes it skips skips define inter define inter define intervals such as shell okay because some sometimes it cannot capture it so we have a large cell for f1 large cell for f1 large cell for f2 or maybe f3 even if some sometimes uh the heterogeneity the course grid the problem with the core cigarette that it doesn't capture heterogeneity with the fine grit the geologist built okay it captured the heterogeneity but the course of grit which is uh required for simulation it doesn't capture so let me say that in uh for the for the uh geologist the the finer the grid okay let's say the cell size can be average uh can can be from uh one uh fro can be two or one uh foot okay and sometimes the maximum reach three feet but for cos grid okay which is preferred simulation sometimes average can be uh six or maybe sometimes nine that depends on the that depends on the behavior of the uh of the reservoir so this is uh so so this is uh how we can as i told you it's agreement the thickness of the cell is agreement between the uh reservoir engineer and geologist none of them can control the the thickness or let's say the average thickness by the way because cells they don't have the same thickness although the cellular grid doesn't have the same uh all the cells have a variable uh thickness and as i told you that is controlled by the heterogeneity and also by the behavior of the production so reservoir and reservoir engineer and geologists need to sit together and agree on average cell size or cell thickness yeah okay okay uh okay host okay i will talk we will talk about how we can determine the context yes sure yes no no there's no there's no difference for both there's no difference actually the dynamic model is the same static model but with dynamic input inside okay so you have you you ha you have the static model okay with all the properties inside the prosthene permeability phases and everything and after you finish you give it to the the to the simulator or the reservoir engine what is the wall engineer he takes the production data and pressure data and fluid data and put it inside the static model that you provided so it becomes a reservoir model or dynamic model then he will run or she will run the simulation and uh check the history match and based on that as we said and based on the uh uh under his history match results they will uh conclude whether the static model is reliable or uh it cannot be used or it needs a modification for uh next phase uh for for development so there's no there's difference in dimension in cell since it's the same cells there's no the there's no there's no difference the same cells we the the the geo model provide to the reservoir engineers of engine just put the dynamic inputs inside okay okay sometimes uh the in advanced stages uh they deal with the what they called partition so uh so sometimes to test the uh because you know you are dealing with uncertainty subsurface is always uncertainty however how much data you get doesn't change the rule you have you always have uncertainty so when uh they tend sometimes to do a partition so they come to the cells like the cell for example okay and do some certain partition just to manipulate or let's say not to manipulate to tune the pressure uh uh along in the in the wheels along around the well and see the impact of the uh uncertainty on the floor behind on the floor or the production of the well uh okay mr taj dean let's have the spray 10 minutes break then go back okay 10 minutes okay okay thank you thank you yes okay i'm back um okay i will continue okay okay we started you kind of started now okay so uh as we said when i encounter when i uh when i want to change okay i want to simplify things okay and change complex to simple what do i do i do i divide it into chunks okay into a chunk a chunks of tasks so so before initiating a cylinder model uh we need to we need to uh apply this principle okay i i have a reservoir okay what's inside this body it has a structure it has a properties it has beds bits of different thickness and different faces uh maybe one or two or maybe um variable deposition environments thickness and extension and everything so we before that we need to divide okay or uh change or from complex simple we need to divide the components of the reservoir into individual parts okay and why we do we do that first to understand to understand each component in terms of quantity and quality and secondly to build a general concept about each part so the first the first and the most important thing to do before even uh building of the model you need to divide the components of the reservoir why to understand it okay so you go to the first component of for example here we have the first component is the reservoir shape and volume uh how is the external envelope shape of the structure the anticline is it a shape of the structure is undeclined is it fall is its salt doom the topography topography or the relief reflected by the type of the contour for both top and the base of the reservoir this is usually provided by the geophysicist so you need to before as a geomodeler before you get into this uh into the model you need to with the rest of the team you you need you you need to do to understand each component separately so that you you can uh understand or you can make a proper interpretation and based on this interpretation you can start building your model you cannot build you cannot build a model just because you cannot build a model for any object just because you are looking at it without understanding what's inside what's the uh what's inside the characteristics of the uh object you cannot model it otherwise the work would be rubbish do we have a complex structure with many forces that divide the reservoir into compartments and also the second compo component is the internal organization the antenna organizations means that the cor such as the correlation uh uh the internal organization is how to the elements inside are related to each other's insightful manner such as such as the degree architecture we said that can be identified by correlation when we put the wills also correlation will by will dwell and also wills to outcrop sometimes you can you can correlate the wills to uh outcrop if there is a nearby uh exposure for the reservoir nearby the field also faces variation in different trends lateral and vertical and their association with petrophysics also the presence of digestive structure microstructure related elements such as drains and barriers so all this uh this is another component that you need to understand that you need to understand and build to use it to build a concept for uh uh uh for for your reservoir before initiating the model so we are still in the characterization phase that we talked about earlier in the year at the beginning of the presentation we were still in the characterization phase second you have uh the uh structure also the structure framework for uh framework which include the falls fractures as we micro fractures uh it's another worth quantifying it's very important component last component is the fluid in place so uh it includes the type of fluids uh type a number of contacts sometimes you don't have uh you don't have one contact sometimes you have compartments if you have compartments this means that you expect to have different uh uh contacts this is encountered in a heavily faulted reservoirs and maybe in like a and also in some [Music] basins like sarawak where you have a different carbonate reservoir cones and each one has a different uh contact uh depth pressure also also pressure and temperature of the reservoir fluids composition aquifer extension energy aquifer energy and boundary and everything so you see all um uh indica as i said to understand to understand object uni you need to uh the complex object you need to divide the components into chunks and you start studying each one separately to come up with understanding so uh what makes the cellular model okay the uh reliable you will hear this uh word a lot during uh when you i'm sure that you hear it a lot in your work uh the word reliable so uh uh i realize let's start with the geological model uh algeological model must be uh must be robust as it takes into account the uh the the the the the static data such as the third or the the false the uh stratigraphic the static the stereographic layers the barriers uh in case you have a multiple reservoirs how much output data is represented it can be uh the degree of representation of this model all the all control the robustness or the reliability of the geological model uh for reliable reservoir models in which which is as we said earlier it's the geological model with dynamic data input it's uh strongly constrained by the geological model and dynamic data and also understanding all uh understanding all this data and make and to be able and this reservoir model should be able to make fluid flow simulation uh that can uh end up with a proper or let's say less uncertain uncertainty uh history match that can be later uh judge okay whether this uh model can continue okay to forecast and uh to forecast and economical evolution evaluation or no okay now uh let's come to distinguish between the uh geological model and reservoir model in terms of data input and type of grid as i said the answer in question for answering a question asked before the break about the grid size the grid is controlled by the cell size actually so uh let's go to the geological uh model it's a static model so it includes all the static data associated with the rock okay and uh in the first degree uh uh it also prefer to be a fine grid okay which is with a small uh scale okay x and y okay for uh nz for the uh for for the cell uh then the reservoir or the dynamic model is the uh it contains all the data associated with the fluids like pressure and protection and so on and the grid size is always preferred to be coarse okay this is an example for fine grid okay this is uh for the fine of uh for fine grid for geological model okay which is used uh for uh volume calculation so what do what are the puts inside we put the uh in each cell we put the uh the the the faces the porosity effective prostate total porosity permeability water saturation gas saturation uh and the net thickness or net gross thickness so all these cells will be filled with geological and petrophysical data uh through upscaling processes okay and mathematical algorithms using different softwares such as betrayal which is the most famous uh modeling platform in the industry now so um so it's the number another thing to control the uh uh similar the okay we'll talk about this in the simulation uh as we as we said that uh in the static model part on geological modeling part is that the grid or the cell size is always preferred to be fine okay why to capture heterogeneity now let's come to the reservoir model or the dynamic or the dynamic model which is used for fluid flow simulation uh so the same inputs okay we will be there for geology uh in addition to uh the scale data including the relative permeability and capillary pressure also the product saturated the saturation and also uh hydrocarbon saturation and also the uh pressure data okay and uh as we said that it's preferred to be a coarser grid so the cell size uh the bigger dimensions okay the the bigger x and y and z so it's the better for uh flexible simulation flow and also there's another constraint okay for a successful fluid flow simulation is the number of cells uh number number of cells actually affect the simulation process itself okay so it's always preferred that simulation will be run in a decent time okay in a time adequate to end up with good results without errors uh usually uh [Music] the reservoirs that i encountered in my uh career uh the uh the cell the the cell number should not exceed me or one million one million or maybe less it's uh preferred can you can you believe that when you have a large uh uh reservoir sometimes in the static model you create the fine grit can have more than three million cells but when you but but when you move from the static model to the dynamic model the reservoir engineer uh do upscaling process up scan process what does mean upscale process is that enlarging the thickness uh the the thickness or the size of the cell so that that can facilitate the simulation of the flow so the number you expected to to be reduced by more than 50 by more than 50 percent uh but this is the this is how it works uh simulation is better in uh coarser grid and that's why uh dynamic model or reservoir model is always preferred uh to be the good can preferred also always to be a course upgrade okay let me move now to the second session which i is have a morning workflow sorry i'm trying to bring it to the uh to screen the next slide just a minute oh just a minute i have to stop sharing okay huh this okay can you see my screen yes mr touchdown okay okay now we come to reservoir characteristic characterization and uh characterization workflow we i intended to put this slide again just to remind us of the steps how we can model the complex things in nature or how we can model the nature we said that we need just to simplify to understand the nature so by dividing it okay into uh to first to understand to understand it characterize it by uh dividing it into uh components okay study it uh unders interpret it make a interpretation then we uh take this interpretation all the information and the synthetic data that we we have got with the input and and put it into into a model and build a reliable static model just to uh to mention in the the beginning of my career uh i uh i passed through a very important lesson i i i want to pass it to you now guys uh when i first started building the first model i uh rushed into the taking the data i'm putting the model uh building and i was happy building the model and everything but after uh after i finish after i finished like my supervisor came to me who was in my position now recently i'm a development geology coordinator uh the previous uh he was the previous development judge coordinator he checked my model for five minutes maybe and he said there's a problem with the data because things is not consistent uh so i realized okay uh that data is represent 50 percent of successful static model so because so data generally okay and most importantly must be reviewed and uh quality controlled okay why because they and this is your job as a geo modeler okay when you are geomoddling you you you need to take care of this point very it's very important when you find irregularities in the data you don't proceed with the modeling you have to stop okay data come from several resources okay as we mentioned earlier you have a data from provided by geophysicist which is seismic data you have data coming from um better physics you have data from core you have data from sedimentologists you have you have you have a basin of input data that you need to pay attention before initiating or starting building building the model because if there's any irregularities okay you will end up with a model that doesn't represent the nature so then it cannot be used for simulation to protect the uh fluid behavior so the objective of that review okay is to decide if the data quality is sufficient to conduct a study sometimes uh okay sometimes uh like okay like an exploration face face okay let's uh suppose that the company drilled one well and they find the good oil okay and they have and they could from the seismic a good seismic showed them that they have a large reservoir okay that have a boundary from uh from from certain uh limit but they still don't have uh enough data to start building or run rcm workflow so what they need to do in this case they will need to uh drill appraisal wells what they call appraisal wells to obtain more data and they choose this locations based uh based on different uh target uh on the to serve different targets it's appraisals that will not produce just appraisals so they will have to take data and that depends on the size and the thickness of the of the reservoir so they need adequate data to start the rcm workflow and to build a model otherwise the model will be very full of high uncertainty that will create a sense of high risk for the company to develop this field and also data review also enable us to plan that acquisition okay for the study purpose let's say that you want to uh drill a development wells okay so you are uh you you you you finished drilling the the the first phase of development the company finished finished was for ten years in this ten years they drilled let's say uh five or six uh wells okay they check the data of these wheels where they adequate uh where they adequate for the next phase of development because the next development now or the next fdp they want to drill let's say another six or seven wells okay do they need to take new data or more data from these wheels and and this is this this is a decision okay for that acquisition this is very important i've been through this process in my career it's very important you need to uh judge review the data and make sure if it's adequate okay or no is it is it complete is it validated is it is there is there a problem with the data and based on that you plan a data you you set a data acquisition which will be part of the fdp or the next fdp so please uh underline this two words data review or team hello yes okay you can see my screen not yet okay let me show it again okay can you see it yeah yeah you can start now okay you're doing me start from the beginning or okay okay i'll start from the beginning okay so that everybody can catch okay okay so uh as uh as we said the um as we mentioned in the first session that we uh that we need to follow three steps to model the nature or to change cam com or to model complex uh things uh the first one is to understand okay so together information uh chunk the components uh divide the components into chunks or the or or isolated so to start to study it uh which will enable you to characterize the reservoir and build a concept then you start building your static model with uh i wanted i want to i want to talk about the importance of that review uh i don't know if you uh uh if you so sorry for interrupt for the sound cut uh i want to talk about the uh that data review that i represent uh the most important part of the whole work of rcm okay rcm workflow you need to review the data either you are the geomodeler who are doing the job either you are or you are physical physicist uh uh in charge of the physical data or the geophysicist who is in charge of seismic data it's the job not only for the geo model the job of everybody so data review is uh very important is very important because it comes from several resources of information it's not uh so you have a data coming from the petrophysicist you have data coming from euphysis you have data from reservoir engineer you have data from all the resources i'm talking here uh as a geo modeler so as your modeler you need to to [Music] to mine this point so you need uh before modeling make sure that your data is is quality controlled and it's consistent uh the objective of data review is to decide if the data quality is sufficient to conduct the study because sometimes uh data you you you have uh let's say that you have ex you drilled an exploration well in a large field and discovered the newly discovered field okay but that is not sufficient how much data do you have you have data from only one will and maybe it's not uh uh validated maybe the seismic is not is not clear maybe it's poor it's poor seismic data so that review will enable you to decide whether you want to drill more wells okay to obtain data or you start or start initiating the development plan but usually in discovered fields they uh they tend especially the large ones they tend to drill what they call it appraisal wells okay for to the job of these appraisals was usually to obtain more data so that the company can understand the uh collect enough information to understand the reservoir and also the objective of that review is to uh is to plan to set that acquisition plan for the study purpose uh let's say that a company made a field development plan and they drilled five wells for 10 years to be able to produce for 10 years after 10 years the company will initiate will do another field development plan by drilling nine wells okay so as as a as a geologist okay uh in charge of the geo model uh you how how do you think do you think that the data okay do you think that the data of the previous fdp wills where enough okay is is enough and it can be helpful for the next fda fdp uh project if it's not uh enough or maybe it's not good quality control then you you will set a plan for data acquisition for the nine worlds that will be drilled in the uh in the second phase uh in the in the second phase of field development plan because that acquisition i've been through this process in my career and it's a very imp very important process data is always 50 of the whole uh project once data is validated and you see it then make sure that your project will be successful and it will goes without uh with with with less uncertainty and minimum errors but uh if data is not quantity controlled and not review reviewed then expect errors and problems and the static model will not you will end up with static model that is not representing the nature or not representing the actual reservoir that you are studying so please underline the word data view or data quality control because it's very important so this is what you see here is the data quality control diagram okay or process okay that should be followed by the geomodeler okay any any geomodeler who was working on before initiating when you up when you upload your data okay you are uploading data from several resources as we said into your m into your model before anything before doing any anything you need to follow you know or go through all these processes check the completeness okay for example do we have a look that covering all the reservoir or not consistency do we have a consistency do do you have consistency between core data and uh we look data for example uh accuracy uh is is is is is there is there a if you have a you are sure that you have uh faces of shale and gamma ray is not is not exceeding let's say 50 or 60 percent in front of shell intervals then here you it will raise uh suspicious about the validity or the accuracy of this data the integrity can you are you able to integrate the data are you able are you able to correlate prostate permeability uh and all all and all these processes you need to go through before initiating the model and make even a report about the this data if they are usable or valid for modeling the reservoir on geological and reservoir level [Music] so uh this is the the workflow summary we come to the characterization workflow uh this stage now after we as we said after we divide we understand understood we understood the components okay we we understood the components of the reservoir okay we are now able to build a conceptual model okay conceptual model is the model the conceptual model can be sketched by hand you can sketch it by hand for uh the city for who by enter for from this logical from the structural data okay from stratigraphic data lithological data okay conceptual model is represent the uh the the base okay or the floor okay from which you start uh doing your walk because you cannot build something that you cannot imagine okay so for to build a structural model you need to have an idea about the structural model how many faults the trend of the faults like cert basin okay or we don't want to talk about separate as a basin but let's say uh like some reservoirs which has the same trend like self basin okay all the faults and set base and all the reservoirs uh all the faulted reservoirs in certain basin are uh most of them not all of them having the same trade which is northwest south east okay so to you we need to uh put this into the code sub concept model so that we can apply it or involve it in the structural model do we have uh also you need to build the concept conceptual model for the lithology distribution sometimes you have a heterogeneity so you have sometimes you have a shady sandstone uh in the point bars area then you have a clean sand in the bottom of the channel after that you have a silt and clay in the uh under the flooding the flooding plane and so on so you you need to have it i'm sure that some of you uh or maybe all of you so uh the the geologists who work on sedimentology for example how they keep sketching drawing the depositional environment distribution in 2d so this is because what he what he's trying to do is trying to build a concept model so that he can provide provided in his report for the company so that the geo modeler can build static model matching what matching the concept so you are making concept and then you build models that match the concept which is uh you think that's what the reality is below or in the uh in the subsurface with some degree of uncertainty because you know it's subsurface there's no 100 percent exactly right so there's always uncertainty after that you come to the modeling modeling workflow and we talked about earlier first is the static model or the geocellular model the new which is uh in which we will put all the static data and the reservoir characteristics from from uh from thickness uh uh uh structural features uh porosity permeability phases and everything then reservoir model where we where we will run simulation and see the flow flow behavior just uh example just to show the comparison between homogeneous and heterogeneous homogeneous is like water okay so it's a liquid has a certain chemical and physical properties okay so it's um some uh some reservoirs like uh cyclins and the stone reservoirs they can have they can they can be considered as a homogeneous reservoirs as we explained explained that in the previous uh session when we uh talked about the channel okay for braided system for example you expect to find parts or all the reservoirs as homogeneous where we have uh similar reservoir test characteristics or reservoir properties but unfortunately in nature it's not always the case it's not always you find the life is easy okay uh especially with the heterogeneous sandstone and carbonate in general because carbonate cannot be uh homogeneous by any means so a heterogeneous reservoirs uh static modeling will be more challenging okay than the uh uh homogeneous uh homogeneous reservoirs uh it will it will take longer time okay it will be there will be a higher degree of uncertainty and that depends actually on as we said on data validity and quantity [Music] now uh so uh uh before uh these are examples for the conceptual model that we will have about each component of the reservoir so before initiating the model you need to have a structure conceptual model okay in your mind you have a concept model how how my uh how my structure model is built okay for example if i have do i have [Music] do i i need to have a conceptual model that shows my structure model goes like this for example if i go to cross section okay i have a fold okay i have a fault how this i have a fall this way okay do i have fault how this fold are trending okay are they going in the same trend this is a cross section okay how many faults also the number of faults is very important to highlight okay uh the contact as well uh the the depth of the the depth of the oil water oil water contact uh if are there fractures are they following the same trend or if i make let me this is the cross section if i go to the plane view okay so this is my this is my uh structural anticline this is my uh faults are they are trending northwest okay are they all trending northwest some of them are they trending in different direction do we have fractures okay do we have fracture corridors or fractures uh uh swarms or fractures associated with steroids there are fractures can be attributed to different uh physical uh processes uh do we have different sets do we have a compartment as my reservoir is one compartment or there is a major or there is a major fault that divide my reservoir into two compartments so this is a compartment and this is another compartment so all this should be uh put inside the uh the model so that it can be representative in terms of structure so this is what you need what you need to have first starting graphic model is the same you you you will use it to show there is war you use to identify the reservoir layering based on or on the information that you have so let's say how many bits you have how many bits the thickness of these bits okay because thickness of the bits is very important to define the uh the grid cells okay and also to determine the modeling parameters you have also the cinematological model as you said earlier the cinematologist after going through core description doing photography or see this and collect samples and see the transition under seats in section under microscope to see the micro faces to see the type of porosity and type of prostate see it for two or three wells then he comes uh with this dermatological model that first that will define what's the depositional environment that i have is it is it a carbon platform is it a for luvia system uh what is it or what is it exactly is it patchy leaf uh is it a bank uh or all this is it is it a shallow is it a shadow tidal flat so all this um uh all this uh information will be provided by symmetrical model and again the system the systematical model that you let's say the static model should have cells okay that reflect exactly the conceptual sedimentological model that the sametologist provided same for fracture and same also for digenesis digenesis is is a key point in carbonate and it's uh and unfortunately it causes a lot of heterogeneity that makes that keep uncertainty level always high in carbonate especially those who are passing specially those reservoirs who pass through complex digestive processes and through burial time then it comes to the also the heterogeneity if you have heterogeneity you uh drilling sedimentologists usually encounter uh shell intervals or let's say high k or let's say called hyper drains associated with sequence boundaries where we have a super k so we call super k or super permeability which is a very high the this intervals or barriers or the high k intervals should be identified and in the input into the uh into the the model as per the conceptual is made then comes the fluid model which is the oil water contact and or gas oil uh contact sometimes you have a complex structure as we said sometimes we have a complex structure associated with the compartmentalization which is associated with faulting or major faulting in the field so you are dividing so it's dividing reservoir into compartments uh and this time you expect to have a different fluid contact and this is what you have to [Music] understand okay uh how these fluids are distributed their extension then you input that into your static model so now uh why we are modeling uh why we are modeling in 3d why we do why do we put put make reservoir in 3d model there are several uh benefits okay for this first is the three 3d is three dimension we are building a model that represents nature in three dimensions so in x and y and z okay also a 3d model can help us to understand or understand or define the heterogeneity distribution and also the connectivity associated with the barriers uh also uh uh the best thing also what what what i like about the 3d model is that they are representing a huge database okay so you have a database uh imagine that you have a database that have uh data from different resources and it's quality controlled and healthy and everything is fine so it's a shortcut for you to use this data to take this data and use it for other studies or another uh projects because 3d model is like a filter okay for uh data so once the so it keeps only the robust and you healthy and quite controlled data and doesn't include the error or bad or an unvalid data also at your static models or 3d models they have data resulted from the integration so when we do data integration we are creating a synthetic data okay which is the data generated from associating data with each other's like their core and picture and will logs together like the uh uh like like uh doing some certain processes like cut off and some statistical equations uh applied on some data that gen from which you generated another data which is considered as synthetic data and this synthetic data is used for uh modeling for for modeling a certain reservoir property or certain characteristic and also and the the most important another one which is important is the volume so we are using 3d to calculate the volume not in wool scale no in the field scale and also the reservoirs or the 3d model uh enable us to uh run okay or do uncertainty analysis okay on each parameter of the reservoir in field and in feed scale in three dimensions x and y and z okay can we have uh time for questions so attendance yes here okay so uh any of the attendants have any question related to the second lecture sorry uh i can't see the questions in the previous uh it seems that it has been lost since the um because of the cut that happened earlier yes but i remember there's a question about the uncertainty and this is what i will answer in the next slide actually because we will go to uncertainty uh later after the volumes so this is this is the question i remember so if anybody has a question please uh type it what are ah this is a good question okay uh what are the qc part okay in upscaling of the data okay when we all right let me bring the whiteboard for geo modeler okay i think i mentioned that earlier but i will say it again so it can be clearer okay so let's say that i have uh uh i have a porosity log okay let me put this well okay this this this is my bull hole okay and this is my uh porosity okay from minus to plus okay i put i draw my porosity now so i have my reservoir here okay in this part okay then this the velocity reduces somewhere below okay and it comes back again high okay so what i have is very high i have reseverance interval so this is my reservoir so i can say that this is my reservoir now uh when you take the porosity log and you put and you do the upscale okay or the scallop inside the 3d model you need to run the upscale okay upscale to capture the heterogeneity okay and the behavior or let's say let's let's be more specific the exact behavior of the porosity log what does it mean you mean you need to put cells in this way oh how how the ups how do you the upscaling process is done is like this let me just finish putting all the cells down here sorry my drawing is not that perfect but i'm trying to make things clear for everybody okay so uh upscaling process is happening uh let's let let's take peter for example but really when you when you tell petrel to run the upscale uh upscale to porosity it comes in this part okay let's and to this cell and see the equivalent curve of porosity okay and see the maximum and the minimum value okay and take the average it calculates the average and it comes to put the value of average okay inside the cell okay and he do the next the next step he does the same so he comes to the curve of porosity along this cell okay and take the maximum okay this is the maximum okay maybe this is the minimum okay and calculate the average put the average inside this cell and keep doing the same all the all the way down what what's the quantity what's the quantity control in this case is that all the cells after finish they should have okay almost they should have almost identical uh behavior as the as the as the original look because this is the input and this is the output so i should have my uh my uh my my upscale porosity going like this okay so for example this is the average let's say this is average i can see that this increase in porosity downward okay so this means that i should have here average 10 10 percent porosity okay here 15 okay okay and here it's uh it's much lower so i say five percent okay and this is increasing so i expect to have um let's say uh 18 or 20 okay so you end up with a grid okay you end up with a grid okay that uh matches the input data if it's not matching the input data then there's a problem so you need to uh re uh upscale risk scale up the cells again and make them smaller and that's why i come i say again that fine grid is preferred to capture heterogeneity if you are you can try it in betrayal or any other software or a miss or or even an excel sheet by the way so the smaller the cell size okay the better capturing for uh for reservoir heterogeneity and porosity and other properties behavior so you have a cells okay which has a values that following the same trend of the input data so we have here average porosity 10 percent in the input data the cell has 10 percent uh average here you have a 15 average then this uh you you you have a 15 inside the cells otherwise it's not it's not capturing if it's not capturing you need to uh scale up again and put smaller divisions for the cells so the output the output the output curve if you are taking the output the output should be like this okay so this is the cell of 10 percent okay now and this is then we have the cell then you have a cell of 15 okay goes like this sorry i will choose another one which it's not helpful so we said this is the cell of 10 percent okay then comes the cell looking for drawing i cannot draw where's the drawing where is drawing okay this is the cell of 10 percent okay and after that you have the cell of 15 percent okay then we come back to the cell of five percent then come back to cell of 18 and so on that's how it goes this is how you control so the cells you you see the trend of the cells it should match exactly the input the input data and and okay don't expect the value to be exactly the same but at least the trend so you have here increase in porosity you the cells should show increase in porosity if there's a reduce one cell or two they should indicate the reduction in porosity okay uh as this presentation will be shared i i i think would be uh i think it will be shared okay what's the approach used for selecting layers and donation of the reservoir okay that depends on the heterogeneity uh if you have as i said if you have as i told as i said your cells must represent the nature must represent each uh component inside your reservoir each phases if you have a barriers okay or if you have uh hydraulic flows or you have a super k or a super k uh intervals associated with sequence boundaries then you uh you you can choose any approach you can divide the uh in the software and the software is like petrel there are several uh scenarios uh to divide the uh to to apply layer and donation but uh you can choose any one of them but that depends on the type of that depends actually on the type of the reservoir and as i told you the heterogeneity yes you mean you can calculate without property i don't understand can calculate directly from static model the old volume and place this is this is defined this is the ultimate target for any static model is to calculate volume in place but if you are talking indirectly okay indirectly you can do it by uh you can you can do it by going back to all the schools just uh calculate the area average thickness of the reservoir and you have the average wattage saturation which and average net growth uh then you and you have a formation formation volume factor you can calculate it but it will be a roughly estimate but static model with by doing the applying the full workflow the static model uh by using by up applying the full workflow you will end up with volume in place more accurate okay reservoir engineers for upscaling uh uh okay there's upscaling process that reservoir engineer do after the end of the static model uh as i told you they don't prefer the fine grid they they like the course grid so there is a process in modeling softwares like betrayal where you can take the whole grid with the whole properties okay and you start dividing the uh using multipliers so you are taking the cells okay and you start multiplying using multipliers so you so you take uh to convert the fine cells okay into larger scales from for example one foot cells to three or four or maybe five uh cells but that's actually uh it will impact the uh the representation of the reservoir and that will cause uh uncertainty to be uh high you you see it's always this is this is very important uh this is very important issue between static model and dynamic model because both of them they prefer uh different grids but several engineers take the they do upscaling for the grid and change it from fine to grid so you have larger cells so that simulation can be in a shorter time and the flow simulation can be also can be smooth okay without any problems without any errors but actually this is not good for the static model because it's no longer represented representing for it doesn't it it's no longer representing the nature or the geological concept that we have built for the reservoir actually it can be done by bot either usual as it uh uh it can be done by by the geologist or by uh reservoir engineer but most of the time is where engineer do that because he's the one who has the pressure data and the uh all the data other times associated with the flow and like pvt and pressure all this data can help him to judge or decide the upscaling parameters that he will put into the dynamic model uh okay uh stu uh okay uh most of the time we use uh stochastic deterministic usually uh uh deterministic can be applied for for homogeneous model but for uh heterogeneous always we use the stochastic uh approaches okay uh uh there are uh s several processes for stochastic because you know it's a subsurface okay um and each time you are drilling a well you end up with a new data and this data somehow shows deviation from the data from other wells which is adjacent and this is called heterogeneity and heterogeneity it means that there is uncertainty so the uncertainty cannot be handled using the deterministic model okay so what you need is the stochastic okay so what we are doing in the in modeling we are uh model each parameter each component which is the phases and there are stochastic algorithms used for uh distributing phases stochastical algorithms for uh continuous data such as the porosity and permeability also for for for and also for water water saturation so uh okay deterministic some deterministic models can be uh built okay and uh and deterministic uh men are using empirical equations such as uh uh high uh the shf saturation height function uh [Music] yes uh the the this is this is one of the famous examples that you can [Music] consider or build it as a domestic model but most of the time especially in carbonate always we are using a stochastic and you you don't run the stochastic process one time for one parameter for one reservoir property you are running you do running uh maybe tens or hundreds of times to take in consideration the probabilistic uh estimation of volumes this is what the within the uncertainty that we will talk about later okay as such as a chef you mean it's a chef yes it's part of the static model actually uh okay uh saturation function uh you've been through different uh papers talking about uh how equal to control the saturation height uh function but the easiest easiest way for me okay for me is that you that you can quality control it by comparing with the plt data okay this is the this is the easiest for me because you know sometimes sw logs or what saturation logs doesn't um especially if you have a micro porosity or you have uh non-connected bugs that cause you a lot of problems uh with the uh uh with the water with water estimation or that or what saturation calculation to generate uh saturation log so it will not be reliable or it will it will be unvalid to use it for calculating hydrocarbon saturation so instead we use the saturation height function okay to build the watch saturation model uh for as i told you to quality quality control it's better to compare it with the the plts to see if there's a consistency or no this is part of the qc can we have no more questions okay can we have a ten minutes break okay okay okay okay okay shall we assume hello [Music] [Music] okay okay fine i think all right now let's come to static or geological modelling workflow okay i took this workflow from betrayal because it's the best it's the best illustrative workflow that you can that you can follow okay and understand it's very uh flexible and easy so we start from the first thing which is the data okay we are here in this part we you put data and edit data input as all available data must be as we said previously that must be qc'd and um and validated before uploading to modeling softwares such as petrol and arm or rms or any other software after that geo modeler okay take the seismic data which include seismic surfaces that represent the statigraphic boundaries of reservoir from top and bottom such as seismic such as horizons and faults used to build the structure model which is the first step or the first uh brick in uh 3d uh static uh model because when you when you want to model this water you need the the the ceiling and the uh and the bottom and this is what this geophysicist will provide you is the size seismic horizon and the top and bottom of the reservoir then you uh after after building after building the structural model okay we we built our structural model [Music] okay so we have uh we we have what do we have okay this is uh our structure okay we have uh our faults okay our faults placed in the we did structure structural models that include also fault uh uh models okay now uh what we will do we will do we will build the cellular model now we will build the skeleton okay the skeleton grid okay and which uh later properties and faces to be filled inside okay so what so so what do we do is that we are transfer building a grid that will have cells goes like this as we saw in different examples earlier so you have cells all the way that should represent and during building the grid we need to mind the trend of the faults because you cannot the trend of the grid should follow the uh trend of the falls especially if defaults they have a certain trend or standard trend north west south east or east west or north north south also you need to mine this when you build the skeleton when you build the skeleton or the cellular model okay uh then after building the skeleton in in patreon it's known as pillar gridding the process is called pillar gridding i will show you some examples later uh next is the uh the strategic statigraphic uh modeling here they say they mentioned that dip dipped conversion okay for uh this for the size of the horizons but i'm showing the picture just to show you the uh macro make uh make horizons process and this process we generate or we build the statigraphic uh model and how we can and the only one tool or the data that we that help us to identify the extension of the bits or the layers is the will correlation when we correlate well to well we can generate layers in 2d then we convert it into a 3d as you can see here okay after that uh now what i have i have the uh structure model the my fault is there the grid there the skeleton is there uh filled with this static architecture is there now we will start processing of cell filling we will fill now the faces faces inside the cells okay uh the the there's uh there's important note to note uh to tell you that the geo modeling softwares are dealing with discrete data and continuous data okay discrete data is like faces it's like faces it has not a continuity in depth okay so you have a three four feet faces one then a half feet uh fascist two then again or maybe uh five six feet uh there's a gap another uh then another uh four three feet faces four and so on so this is this is considered as um as discrete look data continuous log data is is like the wire line logs like the petrophysical looks like porosity like water saturation uh all the data which have a continuous curve continuous curve in this way okay called uh considered as uh continuous log data you you will encounter this uh terms during your work so you you need to mind it so this is called continuous data while faces it have templates like this okay so this is faces one for example another faces two okay this is called a discrete data okay so each template has a certain rock unit with a certain character okay so this is faces one faces two faces four so this is called discrete data and this is called continuous data just to know the difference between both when you are when you guys start working uh in geo modeling new modeling work uh okay so uh faces uh modeling or data depends uh mainly on the availability of core data in the first place if core data are not available then faces modeling cannot be uh accomplished or things will be more complicated for geo modeler to represent the rock units inside the reservoir however faces data is important actually it's important to understand heterogeneity heterogeneity and study connectivity it's also support for reservoir properties distribution in 3d because what makes facials different okay not only the characteristics in terms of thickness and in terms of uh pattern in terms of uh relation to to the adjacent up and down faces no also the each faces has a distinguished uh or distinguished or unique petrophysical uh properties so to understand the distribution of physical properties we need to understand do the distribution of faces so faces it's uh it's important uh let's say that we have built uh this this is this is a picture for gulf x faces model where you can see the channel the channel with the uh the would leave his uh uh would leave his faces on the side the yellow the yellow color uh after that after finishing the fascist model and you are happy with the faces modelling then you come to the physical to do the petrol physical modeling this includes the all reservoir properties so you have a grid for porosity you have a grid for permeability work another grid for net growth another one for water saturation or fluid saturation there are actually there are different resources for physical data that that can be used in the model from lab data to field data which i mean here the wireline data uh now we come to the qc okay uh this is uh can be done by plotting okay by plotting the uh making cross sections okay making cross sections like we do uh like we do like we do in the geo modeling software softwares so what do what do you do in this case let's see for example you want to see the uh the structure so you make a cross section to see the false distribution uh are they identical to the con the concept that seismic provided you uh is the the displacement is it is it identical uh the trend is it is it similar to the concept that you have the facial distribution do we have uh are the faces are following the same uh trend that are they uh are they okay occupying certain uh uh uh let's not occupy uh are they representing the same percentage in reality let's say that for example face reef faces uh according to the symmetries he says that it occupies around 20 of the total faces of the field is it the same in the model are they doing are they 20 or more or less so all these things you need to go you need to consider when you are uh before proceeding to the volume calculation because you you you need to make uh review and quality control for the outputs for the faces for the properties for the structure for everything you can do the quality control by the way after each phase after and this is preferred by the way you can it's preferred to do the quality control on plotting data after each phase after structure after grading after uh statigraphic after phases and after petrophysical uh modeling before you come to the volume calculation do not come do not come to this step okay not come to the volume calculation okay without controlling the data and make sure that you're that you are that your model is representing the reality okay it should it should represent the reality make sure that so make sure that your the average the average uh reservoir properties is similar to the mod similar to the from the model to what you what you have in the concept of what you have according to your calculations uh if there's a difference in deviation do not get into this step okay until you make sure 100 that your reservoir or the model is identical to reality okay then you come to the volume uh volume uh calculation which is the final uh step uh here you assess the hydrocarbon volume in place uh the rock also the not only the hydrocarbon volume place also the rock bulk volume the net volume as well as the poor volume uh by okay by ending by ending the by finalizing this step the model will be prepared to be transferred to a reservoir okay reservoir to start work to to to add the dynamic data and input into the model and run simulation to see whether this model can be can predict can successfully predict fluid behavior and can be used so it can be used to forecast production and field behavior in the future and uh of course the well design it's actually uh at the after the uh step of the dynamic uh part so we usually we design wells as per the uh based on the simulation results okay so the simulation tells us that there are some certain areas where we have let's say that the simulation indicated that there is some certain some certain fee parts of the fields okay let's say that we are in development phase three okay and there are some uh spots of oil okay here we have here we have a spot of oil or pool let's see let's say pool okay and there's another one pool here okay okay here we have a degraded okay we have a degraded oil okay so it's cannot be produced anymore okay and this is another one where we have so what so so so the the simulation shows us that there's a good pool of oil in this uh part of the field so we make uh so the development plan will be based on three or four or certain number of walls to be drilled in this pool to produce and at this in this in this part the geo modeler will make a design or design wheels okay actually it depends on the type of the wheel is it is it trajectory or pilot hole or is it horizontal etc [Music] so this is a very summary for the um workflow of course the the dynamic part it has another uh uh work it has another workflow priority to the wheel design sometimes you can make you can do you can do well well designed just to id just to identify the volume at certain places of the uh field but it's not always the case as i said depends on the type of there is the type of the structure if there's a compartments or no if there's a heterogeneity and so and so on okay uh pay attention to this to this rule you can you will hear it a lot in your job so garbage in equal to garbage out what does it mean it means that we come back to the data if you have a data valid and good and everything is fine so once you are build the static model the simulation will be run and the results will be optimistic and good and the model will be reliable to build to to go for predicting fluid behavior and the develop the field once your data didn't go by the process that you have shown at the beginning at the beginning regarding data quality control and validity then you will end up with rubbish model that will mean nothing for [Music] the project so please pay attention to this to this rule okay i'm showing you here i'm sure i'm showing you here some uh examples for structure model from the gulf x example it's available on the net you can just type gulf x model and you will get a lot of pictures so this is an example for the uh structure model okay where we have this seismic horizon which is a top of up of upper cretaceous then we have the fault model okay this is how we how we model the reservoir then we have uh the skeleton of the cellular grid okay uh usually like the buildings okay you don't have uh uh one roof of one surface of skeleton you will have a multiple okay to support the weight okay and build this in the building in the normal building you build several floors with several skeletons to support the weight of the building but in reservoirs we have we only all we always have three only layers one at the top one in the middle and one at the bottom so and this is and this is you needed to represent the uh fluid you may ask why we need a three because you need to uh to have cells okay all the way from the top to the bottom of the reservoir so that you can uh simulate the flow and represent the reservoir uh properties and this is a stratigraphic flow for the same field so this is what we have this is the input data that we get so we do our correlation we identify where are the channels where are the leaves okay so you have uh you you have a concept about about the distribution so you have a sequence boundary here you can see that layers all the layers are disappearing at this uh uh well just it means that you have a sequence boundary so you take this in consideration when you want to model the when you want to model the dista when you want to build this model architecture so this is a picture for the 3d model of super static graphic architecture now we come to the faces uh i have shown you this in the beautiful picture for uh the mandarin system okay so this is nature we want to transfer it into a 3d 3d model using the core data okay with the help of core data and use we want require to control this data we input into the model and we run and we can see that we have a good representation for uh the channels where we have this we have where we have the channels in orange and uh sided by the levy which is uh the poor reservoir which is the poorest of our properties [Music] which has uh finer sand grains like silt and clay so there is no way to be reservoir but it can be a good ceiling potential [Music] did we come to the properties model oh there's a missing picture here picture sorry there's a missing picture here let me just return okay that's it uh okay we come to a property model uh for for property model prostate mobility and and what saturation net to growth what we use the most important data is the core analysis we need uh there are two types of uh core analysis carried out to model a reservoir to be uh more robu to be more robust and reliable and more representative to properties hence can be used to predict flow behavior first one is conventional analysis what what the samurai abbreviated as a ccl okay uh so we have the porosity okay of course the variety of cool we have the grain density and the bulk density all the four are considered are input into the static uh model in addition to reservoir characteristics such as thickness and as well as faces so these are the most important for four core analysis that we need to input into the static model to build reliable reservoir properties uh model you know porosity is the first property and it's it's the most important property that we uh from which or based on which the other properties will be built like permeability and water saturation so it's very important to be quality controlled and to make sure that there's no problem uh with it under our it's considered as continuous data so it can be distributed in 3d using several algorithms but the most famous one is the sequential gaussian as sequential gaussian simulation which is the famous one used for distributing porosity reciprocity then we have the special core analysis unfortunately special analysis is not always obtained like conventional analysis actually depends on the type of the will that will be drilled it also depends on the heterogeneity of of the reservoir and the target of the well so what we have we have the relative permeability okay we have the capillary pressure and water or water ability for for capillary pressure it's the data that can be uh capital pressure which is the uh the the the difference it can be the the difference in capillary pressure or capillary medium between the wetting fluid and non-wetting fluid between the rock and the fluid uh usually it's very important data to uh for for the static model it's very important to because it can help to identify the transition zone so when you have a capillary pressure data so the same as we have but this y-axis and this is x-axis okay i have here the watch saturation okay [Music] and here i have the pressure and this is my curve [Music] this is identical reservoir a good or good reservoir curve sorry it's not that perfect drawing but i'm trying to make things clear okay so uh what's the benefit from the capillary pressure first we can use it to model watch saturation and by transferring by creating shf or saturation height function by using saturation height function and create the saturation saturation height model that we talked about earlier it also can help us to identify the transition zone okay transitions on for [Music] it it can be measured from the point where we have irreducible watt saturation it means that no more water can be expelled out of the sample okay so this is my irreducible what's uh reducible watch situation okay and the point from uh and the point from which uh what's what situation start to uh uh decrease from here so this is my transition zone so it can help to identify the identify transition zone so that we can put it into the model as well before providing to simulation so actually the curve behavior is the different from sample to another for for example in very perfect reservoirs you will find this curve move inside sandstone sometimes even the transition zone you cannot find it or sometimes it will be very minimum but in poor reservoirs the curve will extend outside like this and so i saw some curves in carbonate goes like this so it's so much and the uh and the average saturation was 90 uh percent and the permeability was very bad so it wasn't a good representation for uh it was taken from poor reservoir property so the the cap whenever the capillary pressure curves moves uh inward it will indicate better reservoir properties outward it will indicate uh bad reservoir properties so there's no need to so it does not need to model it because it will be 100 or near 100 water [Music] now come to uh uh the kf integration or let's say let's say one of the most important integration process between porosity and permeability uh once we are once we have got the porosity model okay we want to model permeability okay and how to do that how we can model uh pre permeability uh until now okay uh in carbonate i'm talking about in carbonate still there's a problem with the calculating or generating permeability logs for carbonate there is always some degree of uncertainty that cannot that cannot enable the geomodeler to be sure that this data can help can help to build a reliable uh permeability model so until now what we do in carbonate we until we do we do integration process between porosity core and permeability core okay and we make a plot uh usually okay if we have a relationship between prostate and permeability then we can indicate it by a linear regression okay or the linear relationship as you can see here okay from from the scale we can see that the more the the change in porosity is followed by a change in permeability so i can see that there's a good relationship between both since there's a good relationship between both i can model the permeability okay based on the porosity value and this is better and this is better for me and we believe we tested this on some models and we found out that this way is better than depending on permeability logs where you will have some degree of uncertainty uh it can re it can be maximum at some intervals of the reservoir so always in carbonate so far so far use this method so uh this is the experimental correlation okay that we that you will input into the model okay you will put into the into the model so you you have the log k which is mobility because the variability in log scale okay and you have the a which is the slope of the of the curve okay and we have the porosity value or the plastic grid that you have built plus the intersect value which is taken also from the curve okay exit can help can help you so much with that okay just a minute okay so uh uh yes so so this is the plot or the famous plot that used in the heterogeneous heterogeneous reservoirs to predict variability and [Music] model okay now this is an example for girl fix okay we have some pictures here for girl fix so we have a we have a uh porosity uh model we have prostitute we have a permeability cellular grade we have net to gross okay to gross uh grit and we have finally water situation what uh what saturation uh grid uh earlier someone asked me about the uh i just remembered now that uh someone asked about how we identify the oil water contact uh actually there are two ways okay the better physicist can identify the uh oil water contact okay which uh the old school okay which is the will look uh interpretation so we have and resistivity logs resistivity logs and or watch saturation either from resistivity okay or from water saturation log or and the most on the most accurate one okay the most accurate one is what we call it the pressure gradient which is uh indicated by xpt which is a tool okay experti is a tool that measure or or measure the the fluids pressure at different intervals along the borehole okay so we end up okay as you know in the poor hole what do we have we have two fluids okay and standard we have two fluids or three three fluids gas oil and water okay so when the when this tool xpt takes the measure the values of pressure at some intervals for the fluid okay and this is the the depth or maybe oh sorry this is um it's vice versa this is the dipped okay this is md okay the measure dipped and this is the pressure so when when we plot the values of the fluid of the uh fluid uh different fluid pressure uh you you we will we will find a change in the trend okay when you find the channel you what we do is that when you find different in the trend let's say that this is a measurement for for for interval with water okay and there's a change in the gradient this is a certain gradient for the part of the reservoir where there is water okay for example then we have we have a change in the gradient okay that becomes something like this okay so at this interval okay so this is the oil and this is water so what the better phases do actually it's a reservoir engineer okay it comes exactly at the change in the behavior okay at the point where we have a change in the in the trend of the gradient okay and take this as the contact between oil and water and this is believe me the most accurate way better than depending on the resistivity and what saturation because as i told because you know sw is calculated from resistivity okay and resistivity depends uh is um is affected by borehole conditions and reservoir conditions and reservoir conditions as well so this is xpt is more reliable to distinguish or identify the fluid contacts for for which you will build another grid to be used by the reservoir engineer to simulate to simulate the flow then we come to the volume assessments these are we all know this is famous equations for calculating stop okay we have also other volumes to be calculated we have the net rock bank volume we have the metabolic volume we have poor volume okay which is the volume multiplied by porosity then we have the hydrocarbon pool volume for oil only we have hydrocarbon proof volume for gas the stoic okay which is the standard oil in place and initial gas in place all these volumes are calculated by uh the geo modeler okay most of the soft all the software is actually in the industry now like betrayal like are misled they provide you with all these volumes this is an example for uh well design okay so you can see this this is what we are talking about the integration of the data so we have here cross section for the faces you are trying to capture or follow the good phases where you have a good oil okay you also show the seismic data you see when you are working on as a geomodeler you need to be artist with data visualization as well so when you are visualizing data okay in one window like seismic like faces like porosity like everything so that it can help you to choose the best location okay in terms of structural features in terms of petrophysical properties and you design your will to follow the best reservoir properties away from troublesome structures like false and or or fractures okay before i proceed this one i need to explain something regarding the uh regarding the uncertainty because uh i remember that uh someone asked me about the uh so sorry the the the the the presentation cut uh that happened uh earlier uh uh to cause some all the questions to decide to disappear it look looks like zoom restart when when it restarts it deletes all the messages so you know that what are the inputs into the static model let's count them together it's let's say sw okay we have the porosity we have the permeability what else we have the need to gross and also we have the faces okay okay uh each time okay when we run the volume okay we are uh running the volume based on data which represent samples from the field you when you have a field okay and you have wells this is a will this is another will this is another wheel this is another wheel and this is another wheel okay what do we have is samples okay these wells are just samples from the whole reservoirs okay and we are building a volume and assist the volume based on this data so this means that we are not calculating the original or the exact volumes in the subsurface why because we cannot we cannot drill well each one foot or each one meter to to uh calculate the uh to calculate the exact volume there's no way it can be this can be done so what we are we are doing we are depending we are calculating our volumes based on what's called samples okay from the reservoir so these samples are data okay so each time we use this we are taking this uh data from different will samples okay so sw so let's see let's say for example porosity okay we have the average porosity average porosity from this well is not it might be not the same as average velocity from this well okay so i expect deviation in the data okay and this deviation in the data should be considered when we are calculating the volume and this is goes in this part we do what's called uncertainty analysis for each reservoir parameters for porosity for permeability for net growth for sw and run probabilistic analysis okay to see the impact of each parameter okay on the volume and we end up with a probability with the problem with probabilistic volumes which is the 1p 2p or the probabilistic cases we call the provision cases and three p so p one p is the proven okay which is the uh which is the uh which is the volume that we that ninety percent will be produced okay two p which is the fifty percent which is the fifty percent probability that the uh uh the calculated volume will be produced three p it's the uh it's the it's the less case or let's say let's say the the case uh um that list list 90 90 that this volume will not be produced so we are taking consideration all these uh parameters all these parameters to uh as as for uncertainty analysis to end up with three uh probabilistic volumes one volume for one p which is the proven we have the uh proven uh probable and we have a proven probable and possible that's why you have a one p which stands for proven which is approved which is i'm sure that it's there and i can produce 2p it's the proven probable okay it's 50 percent okay to be produ to be produced 3p it's ten percent to be produced or 90 percent that it will not be produced so it stands for proven probable and possible so i hope this uh uh will be understood but by the way the answer to the analysis is uh it's it's a big part of rcm and it's a very long uh process and it's uh you and it's using g-steps and it uses g-statistical uh uh processes so it needs a geologist or geo-modeler with good knowledge of geostatistics so to run to calculate the three cases of the of the volumes by considering the uncertainty of each reservoir parameters even size even seismic structure structure and false are considered into this into the answer into the answer into the uncertainty of the volumes um okay so now uh i will talk about the final part of our session today which is the iteration process between static and dynamic dynamic uh this attention process uh this iteration process it's like a repetition okay so this the gmo the geomodeler when he finish building the static model and the and he's happy with all the inputs and the outputs that he ended up with he passed the static the static model to reservoir to to reservoir let's say that we are in this phase of static model when he finishes he provide the static model to the reservoir engineer to do scale up what we said about is that it's transferring the fine grid into a coarse grid then after that the reservoir engineer takes the dynamic data from pressure to protection to relative permeability and put it into the dynamic model and run simulation okay and after that he checked the history match so history match when it comes to it uh when it comes to history match we have history match for uh uh for uh the quant for quantity of oil we have history match for quantity of for pressure we have history match also for gas oil ratio and for water cut let's say that we have okay we have a production data for oil for quantity of oil okay this is the quantity of oil okay this is the the quantity of oil okay and this is time okay let's say that the field has been producing since 2000 okay to up to 2020. okay and we have a community protection goes through time like this okay at 22 this is the actual okay this is the the actual this is what the field gave us through this time through the the last 20 years okay if if the simulation okay predicted the history match that respect this one let's say that history match the predicted history match that the dynamic or the simulation predicted goes something like this okay you don't expect it to be exactly the same but at least to be the same trend okay then at this point we can the reservoir and the geologist okay they can agree that the static model is representative okay for the uh reservoir and can be used further okay to field development project and to forecast and continue drilling but in case the hist the history the history match was not like this let's assume that we run the simulation and the reservoir engineer ended up with something like this which is much lower in terms of quantity of oil compared to the actual okay at this point the the what will happen the static model will be uh sent back to the geo modeler to work on the answer or to work on the prop on the problem because this problem uh actually it's not a problem it's uncertainty so that this means there is uh some uh reservoir parameters that need to be checked and this is not done by the geo model alone this can be done with support from the reservoir engineer and the petrophysicist as well okay so it's uh three they are the the three person these three expertise work on solving this problem petrophysis is involved when when the problem is associated with prostate permeability or or geologist if this problem is associated with the uh fault location or the the presence of fractures or the absence or the existence of uh strange faces so so so this iteration process keeps going until the uh the history match is uh perfect then the as we said the static model is really considered to be reliable and to be used for forecast and to locate uh the future wells that will produce more oil in the future okay for me it's the end of the uh this is the end of uh session so let's come to the questions so let me go back to the uh candles of our engineer static model actually it's not easy because you know where the static model is geological domain you cannot use it's uh there is the four engineers uh spent uh five years studying the uh dynamic uh the dynamic properties okay the pressure and production so it's he's mainly concerned in the in the in the fluids yes in college they give courses about geology but it's not but just for knowledge it's not for practice but the geologist is the one who practice practice geology with some background in the reservoir engineer as well but for a static model it's better to be carried out by uh by a geologist unless unless you are interested in static model you need to practice you you you need to practice or get more exercises in static uh not in static model in geology because this geo modeler is multidisciplinary man you need to be familiar with the sedimentology you need to be familiar with petrophysics you need to be familiar with structural geology all this take years to understand because by practicing only you can accomplish good work not only by reading articles or books so uh for me i i think static model is for uh or geologists is the best one for a static model dynamic simulation of course it's for reservoir engineer geologists cannot run simulations because as it has something to do with the dynamic parameters especially pressure and fluids what is the tolerance of volume between static and dynamic do you mean the the consistency if you if you if you are meaning the consistency in just a minute please sorry uh for uh for if you are talking about the consistency actually there should be a consistency between static and dynamic uh volumes uh otherwise there will be an issue with the uh water saturation because uh uh uh some sometimes uh they have reservoir engineers they calculate what saturation using different methods they don't depend uh on what saturation uh by using uh will logs or capillary pressure especially reservoirs with especially with reservoir with high aquifer energy so expect in this case some inconsistency in terms of volumes but both of them should work on on fixing this issue because you cannot go ahead with the simulation uh as long as you have difference in or inconsistency in volumes between static and dynamic uh okay about tell us about the common pitfalls during the petrophysic integration with the sedimentary model actually uh sorry i didn't have enough time because uh three hours we didn't have much time or talking about rock types but maybe in the future we can have uh sessions about rock types rock types is the uh is the challenge work for petrophysis is this it's the work where he is trying to integrate faces and uh reservoir properties in one uh rock in one rock type so this means that you have uh what's what's the uh what petrophysis try to do with the rock type petrol physicist try to have uh to create a rock okay rock unit okay this is a rock unit that has okay a unique faces okay unique faces let's say that we have a unique faces like let's say a grain stone okay for uh carbonate or grain stone or pakistan faces okay it should have this this unique faces okay should have a unique petrophysical properties like a unique porosity or average unique porous average porosity or unique range of porosity unique range or average permeability same for water saturation and same for watch saturation same for poor volume okay and average poor volume and also average pc curve as well as average uh relative permeability this is the most uh challenging part for petrophysis to do to generate rock types okay uh that combine all reservoir properties within certain um uh woodens inside one certain uh unit okay so this is we call it truck type one okay rock type one rock type two it has another certain uh or certain uh reservoir properties range or or average believe me it's not easy task in carbonate and sandstone it can be easy but in carbonate especially very very heterogeneous carbonate i uh uh um i work and i'm still working on some carbon reservoir which believe me uh rock rock type is still a nightmare for uh to accomplish but it's very interesting part so this is how you can uh uh so just this is one of the pitfalls of the uh i i don't i don't want to say it's but fall you can see that it's a challenge okay it's a challenge to accomplish this uh because you cannot you you cannot provide the uh you cannot provide the geo modeler uh reservoir properties uh to distributed without a guidance you need either fishes or rock types and most of the cases in carbonate you need rock types not faces because you know carbonate faces you might have uh pakistan fishes with the good and bad reservoir both good and poor reservoir properties so you need the rock type to represent all this okay yes okay discrete okay this okay discrete data the discrete as i said if you uh if you uh if you remember we talked about the security data it which which is not uh it has not continuity okay uh like logging data logging data is why it's continuous because it as the tool goes down along the borehole so this is my tool so this tool keeps this tool as it keeps going down okay it keeps continuous it's going to provide us with continuous data okay so it's uh each uh half a feet it uh it record number it gives a number so it gives us a log looks like this this is continuous data okay but for core okay when you take a core okay core is not continuous data because sometimes you are taking it taking a curve from separate intervals so even the the the the phases inside is not a country or continuous so you have the uh okay let me take some shapes okay so you have you you have uh you you have faces okay let's call it faces one okay which extend extend from let's say from 100 feet let's say to 105. okay then we have another different faces okay it's not the same faces it's it's different faces okay extend from 105 to from 105 to let's say uh 110 okay then it moves again sometimes you have a gap you you you don't have even faces due to the broken rock samples or damage so that's why it's not uh it's not because there's no continuous recording okay so you have a face as one a fascist one is continuing along the the borehole from top to bottom you can see it's continuous data but this is not this is not possible you have a different faces so you can consider it as discrete data another example for discrete data is fractured data fault the the false is discrete data fracture fractions are also the discrete data the barriers is discrete data as well it's not continuous only one leg you can say that only when logs is can be considered as continuous data this simple curve yes yes it's simple yeah you're right for only homogeneous rock yes in reality it's different to from this of course yes it's uh in in carbonate sometimes the same facials okay some faces the same phases you find different behavior for the same phases they can tell you you are taking the same phases okay from different and different depths when you take samples for scale data and you see the pc curve you see different behavior some of them is good indicate good reservoir and others uh they are not they are poor reservoir but it's the same faces of course it's not easy i'm just showing the uh simple case but believe me it's another challenge it's not that easy can we predict sand protection based on reservoir us it can be predicted okay uh actually uh uh okay let me think about it can you predict sand protection you you you you you are talking about it i said you are talking about this uh the net pay and in the new wells or it can be predicted i'm not sure if i understand your question exactly uh i'm confused if you are talking about in terms of protection or you are talking about of net pay and the relation between depth and pressure should be in dvd or uh should be should be in dvd okay it's okay it's up to you if it's a vertical well i don't think that there will be so much difference between dvd and md but if it's in horizontal well it's it's preferred to use dvd yes [Music] okay uh could you please tell us about your experience okay so then the major largest sand is available built into the static model enter the diamond but what if we find minus moles recovery factor now okay we are talking about small scale reservoirs if we find one structure okay had major and minor reservoir you mean one structure has a major and minor reservoir together okay in case uh you okay you will go through the through the same workflow if you have uh either uh okay actually what controls what what controls what controls the the development of any uh of any reservoir uh is the volume is it economical uh okay the structure can be large but sometimes it's large but it doesn't have uh uh economical uh volume to be extracted so what's the use from uh from paying for from paying money to uh exploit this reservoir so but but since it's economical that this means that there is a business so you can you can go through the through the same process or the same workflow of rcm yes okay but if we find um you know so if if there is a if this is a good quantity of hydrocarbon to be exploited then you can go through the same workflow if there is no then why you should even think about it proxy models no do you know about the proxy model for the design estimate recovery factor i i think this is the reservoir part you know unfortunately i heard about it but uh this may there is a war part uh yeah i'm not sure if uh i don't think that i can answer this part do you know about the proxy model or the oe design of experiment to estimate the recovery factor uh in percent for the minor reservoirs that were not modeled in eclipse if yes can you explain no haha sorry i i let this part for uh one of the attendees who is reservoir engineering maybe he can answer this part for me i'm jealous so i cannot sorry okay how we can do faces modeling okay as i as i have shown you earlier in the slide let me go back to the slide okay this is uh faces okay face is modeling when you when you want to do phases modeling first you need the data okay so you take uh so you take the core uh the faces for from the sedimentology sedimentologist do cool description okay he protect he described the core uh plot make a plot for core logging that shows the the fascist distribution of faces from bottom uh to top then he provide what is what he does he provide the this faces as a digital data okay for the geomodel to input into the static model for each will after that we go to scala process okay scalar process enable us to input these phases okay into the cells okay so let's say that you have for example we have the uh let's say that we have here uh okay we have let me okay this is phases okay and this is another one this is another one upscale process enable us to do what to back okay this is faces two this is faces three okay this is the road that this is this is my raw data along the way this is my well and this is my face this is my face now i i do scala process scalar processes what do the software do is that we'll change these faces into the cells that i have i have a cells okay like this sorry again drawing is not that good but i'm trying to make things clearer okay so what it does it will come to the to this phases okay measure the thickness then come to this cell and fill it with f1 okay now second cell we are it's still adjacent to the raw data of faces f1 okay so it gives f1 after that in this cell what it does it comes to see the the software see that there's a boundary between the faces one and faces two so which faces to put so in the upscale upscale process you can define the way of upscaling okay based on arithmetic value uh the maximum value or the minimum or the minimum value or the most of okay must of approach most of approach when you choose most of approach the software will it will see how much of this faces is how much this how much this phase is occupying relative to the cell beside it okay and how much uh the space the second phase is occupying let's say that from here to here okay it's about seventy percent okay seventy percent of f of faces f1 occupying this uh cell and and 30 will be occupied by what by the second phases this this faces so what does the software will do it will consider that this cell is for faces one uh one since it's representing 70 percent the cell representing 70 percent of the space of f1 so it will consider it as f1 then after that okay it will go to the next cell it's in f2 it will give it f2 and so on after that you will you will you you will distribute the faces using stochastic algorithms okay like uh uh indicator that there are two famous uh uh algorithms which is the first one is truncated gaussian simulation and indicator gaussian simulation the most famous uh stochastic algorithms that's used for distributing phases in 3d so i hope this answered your question so any other questions what if we dave vlogs only you you mean use logs uh uh only okay i you you mean in case you don't have faces okay all right in case you have no faces and this is not uh favorite because fishes need to be uh you you need to be applied so that you can build the uh it's essential for building a proper property model then you will be using uh you you will start doing uh what we call it uh petrophysical group uh using uh cut off cut off values because what's important about faces is that we need to we needed to distinguish good reservoir good reservoir faces from pores of our faces we know that pools of our faces like shell it's not uh contributing to protection because they have no perm they have no good permeability this is what we need about faces if there's no phases then we will have to follow another approach okay using a cut-off result cut off and apply it on other reservoir properties like porosity to distinguish where we have good reservoir good good reservoir good reservoir rock units and which have good reservoir properties and where are the pores of our property or pools of our faces which have uh poor reservoir property because this is very important for the simulation so this is in case you don't have uh uh in this case uh you need to have uh faces because you need to know where are the poor reservoir properties located and good reservoir predicated because it will help a lot for simulation and to predict the flow behavior okay for okay for uh value gram okay for uh all right for value but okay for uh value gram uh uh the variogram it's uh special analysis polygram is generally special analysis you it's so used to to predict the distribution of the faces between the wells okay so let's say that you have okay let's say that you have uh well here okay and you have a will here okay you have a faces you have faces when you you at some at certain layer you have a faces one here okay okay and you have a facials two here okay you don't have faces one okay you have faces two at the same depth oh let's uh so what does what does that mean this means that phase is there's a change in the distribution of faces from phases one to faces two let's say that there's about one kilometer or two kilometers from this well to this well okay this is one kilometer okay this means that faces one just somehow some some and some in some parts between two wheels disappeared variogram help us okay to uh mo to map okay or model okay the distribution of phases one considering the reality that it's not uh extending beyond faces too so uh the the variogram the the very gram we will consider the the the direction or the dimension in x and y direction okay and you put it into the stack into the stochastic process and run several cases okay and each time you you expect the the simulation will give you a sense of similar simulation it will give you uh the uh the uh the the boundary or the limit between phases one and faces two maybe sometimes give you here sometimes it gives it here sometimes it gives it here it's based on what it's based on the experimental uh values the this is called the experimental diagram based on the major and the minor or the x and y values that you input then you compare that to the new comp how you can how you can make sure which one is best is by comparing with the input data that you have and see the percentage are they identical or not okay to predict the value of primitive core is not provided well uh actually yes i told you it can be uh done but as i told you in carbonate i'm talking about carbonate reservoirs it's not easy task several papers i read several papers talking about the efforts to predict predictability where logs but some cases were successful okay especially in [Music] in in carbonated reservoirs where you have some less and less heterogeneity compared to uh complicated uh reservoirs but it's not always uh but it's not always applicable to other another reservoirs some of some of these methods were empirical okay using different equations like timor and texar and uh others uh there is also you can you can use multivariate analysis by using integration of several logs but also this is not very successful and in carbonate so i think cool data is mandatory for probability prediction in case of carbonate okay any more questions okay it seems that we no longer okay can be of course yes it's not only about faces modeling only uh the more the more data you have the less uncertainty you have in in any reservoir parameter so if you want to do a good phases modeling you need to have enough uh or enough samples uh from the reservoir to represent faces to represent phases uh to represent the faces of the reservoir if you so so so let's say that if you have uh uh a large uh large field with the three four uh wells with the cool data and faces of course there will be high uncertainty for the distribution of these uh faces but let's say that you have a the 10 or 15 uh core data from these 15 uh well make sure that uncertainty will be uh reduced uh and the faces on your faces model will be more representative for uh the reservoir fishes so uh always more data more data equal less uncertainty you're welcome you're welcome okay okay uh uh i think it's done for uh for today i think that no more questions yes for your effort i think that it was very very interesting uh workshop for us and also i think that all the participants enjoyed the discussion with you especially regarding your great experience regarding the aesthetic modeling and drug typing and so on so thank you for your time thank you thank you very much thank you thank you everybody thank you uh and also thank you for all the participants that our switching until the end also i will i wanted to remind you that if you wanted to have a certificate for attend this for that workshop you should fill into the form that its link was sent into the zoom chat also if you wanted to if you wanted to enroll into our upcoming course regarding reservoir characterization using core data also you can register and view the course content from the links that will be sent into the zoom chat thanks everyone and hope to hope to see you into the next event thank you okay thank you thank you you