Oil&Gas Projects

Assessment, management and development

Subsurface part

Подземная часть



«Subsurface part» is the most important stream in the work of project team. If you do not get right estimates of reserves, number of wells, flow rates and drilling schedule all next results of the project will have large uncertainties in possible values of profit. It is a big problem and headache for investor as known.




For the project team the first problem is right estimation of volume profitably recoverable reserves. Under the profitable recoverable reserves we understand next: how much reserves we should (and can) recover to make project profitable.  Definitely one may ask – «What will happen with the recovery factor?» In this case we are talking about the efficiency of recovery technologies at present time only. In other hand the production profile is not the aim of our interest but the most interesting part for us is the economical efficiency of standalone well at 5-17 years horizon. A project references, recovery factors written for more longer periods of the time but 95% of project profitability depends of this time horizon. 


In many real cases the volume of profitably recoverable reserves is essentially different from the state balance. Even if we not get examples with multi layers field, where we should make the complex development plan, in case with one layer we should define risks for reserves of C2 rank and for resources C3-D ranks. In addition, as part of project work, geologists need to prove not only the volume of reserves and resources from the perspective of the likelihood of their presence, but also take into account the qualitative characteristics of reserves - the ability to extract them to the surface by conventional techniques in the industry, and then making a profit from their sales.


A joint work of geologists, development engineers and economists is required from the first stage of the project to define the cut-off value of profitable volume of reserves. The iterative approach is commonly used for this aim. Our experience shows that «The first step of reserves estimation is substantiating of methodology which is used for estimation».


What is our base? In many cases, we do not have enough volume of geology information. This is why the geologists have been making researches for brownfields. It’s also known that the geologists are short of the information about the field at the every stage of field development.  Really, it is a right way for the researchers who wants to get the full knowledge about the object, but as for the project this strategy is incorrect because the high volume of investigations requires high investments and time. As a result we get a delayed production and significant money discount factor. One can say that without these investigations we cannot guarantee a profit for investors. Really, we cannot, but if you answer so, you are not a project manager. The key project efficiency lays between risks, investments and time balance – we spend money now but result we can get some years later.


Our opinion there is a compromise: we should define current range of uncertainties in main volumetric data and identify parameters which range of uncertainties is too wide. What does “too wide” mean? On economic language, it means that edge points of this range gives both negative and positive values of NPV at one time. For shareholders it means that project can yield both profit and loss with the equal probability. What do you think the shareholders solution would be in this case? Right, they ask – «How much money I need to spend to reduce the range of uncertainties and to move it to the positive or negative zone of NPV scale mostly?


The aim of the project team as we showed above is to find out the most powerful parameters to reduce uncertainties.  At the first stage of investigation, we should find an answer on question – “Does the project have perspectives with present conditions?” At the second stage, we have to find an answer on question – “Can we increase the effectiveness of this project?” So, according to these questions we should not provide all of possible investigations at one time. For every step of the project we should have clear view in what parameter to study and it influence on reducing geological uncertainties: from the reserves estimation to production forecasting and identification of technological emergencies during production.  Our experience shows that key parameters have been changing from one field to another. Certainly, we wanted to develop some templates, e.g., one - for West Siberia, second - for Timano-Pechora, the third one - for East Siberia and to implement these in the future with a small amount of time. Sorry, the reality is not so simple. Why? Оne can imagine that the task to find the field-analogue is the same as to find two absolutely identical men.




Questions about core sampling and corelab are not so simple and relate to the project necessities.  For example, core of Cenomanian sand layers in West Siberia that contains information about porosity and permeability of common collector is poorly cemented and is being crumbled in the hands. As a result we cannot get a required information for dynamic modelling. As for low permeable samples we can get them from the well but we don't need a lot. Should we get it from the well, should we pay money and time? Sometimes we should. E.g. to develop technology for direct drilling.


«Subsurface part» is the most important stream in the work of project team. If you do not get right estimates of reserves, number of wells, flow rates and drilling schedule all next results of the project will have large uncertainties in possible values of profit. It is a big problem and headache for investor as known.


 Another example is core of East Siberian fields. We known that oil and gas is mostly filtered through the fractures. Permeability of these fractures defines rates and production of wells. Situation with core samples is the same as Cenomanian, core samples are being crumbled after extraction or it properties has been critically changed after decompressing. As a result we got core samples with properties that differs from collector properties in situ.


We have positive examples at the same time – development of Jurassic sediments of West Siberia (Bazhenov, Abalakskaya, Tyumenskaya and Sherkalinskaya formations). No one can get an effective result in formation developing  without a significant volume of core study. Or one can get a casino-effect in production. By the way, these formations has good basis for qualitative core sampling and next researches. 






Poorly cemented core sample (sand)





Core samples from high permeable fractured rock formation 


Many people from the oil & gas industry knows about fracking technologies for low permeable reservoirs of Bazhenov and Tyumenskaya formations in West Siberia. The question is - who wants to spend money for core geomechanical investigations and integrating results into dynamic model of the field? It demands rather long time and relatively expensive if to compare with other core investigations. For your information (based on real experience) saving 0.5-5 mln rubles can yield a loss of 100-400 mln rubles (dry well cost) as the result.


There is a possibility to investigate in old core samples study. One can get it  from the State Funds. Note, you are to keep in mind that time is good factor for cognac but not for core. By the other side, some key investigations such as sedimentological and paleontological analysis are possible to make on. These investigations are important to obtain clear geological understanding. The routine investigations we are recommend to carry out with old core samples only in case you have not enough money or time to drill a new well with core sampling or you do not have enough time to reach a decision.


Logging investigations


Logging has the same situation as coring. The large amount of wells has classic set of logs (Gamma log, Side-Wall Resistivity log, SP log) and modern logging technologies, which allow to get rock properties with more detailed resolution. What we have? If we will do interpretation and recalculation into known rock properties (thickness, porosity, stratification factor) then in one case we would get rather smoothing result and in other case we get very detailed result. If we try to compare them in one observation point then we will see completely different pictures.  Indeed, these pictures will be similar but not more. 




Different methods of logging (Russian and Western) 


What we should do? At first, we can use some known methodologies like statistical data processing, neural networks and some other. Basing on these methodologies and some wells with old and new logging methods and qualitative core samples, we can find common factors and solve task for all wells using inversion approach. One should go from the general to the special, from the smoothed to the detailed.  In the second, one should keep in mind that information we got from the logging data has a local character. Therefore, if we have log data with high resolution at the point (where the well is) we cannot extrapolate this one out the well radius. Otherwise, if we would find common relations between some geophysical data we can use it for stochastic. As a result, we can describe well spacing more qualitative. Our conclusion is that number of these modern technologies should be determined  according to the object heterogeneity. For the terrigenous deposits of The West Siberia 1-2 wells with extended set of logs are enough for minor field. But for The East Siberia 1-2 wells are required for one tectonic block.


Exploration seismology


In this chapter we will talk about the process of acquisition information to outline  field’s boundaries. The main tool of this is an exploration seismology. People who involved in this area know that result of investigations is often multiple valued (now 2D seismic is widely used and 3D seismic is used often and often) and nobody knows what to do on the next step – should we do add seismic profiles or to drill a new well and get direct information about formation and field at whole. 


It happens that existing geological and geophysical data has a low quality (results of interpretation 2D and 3D seismic, drilling results and well tests). That is why we should estimate practicability of additional number of works (reprocessing and reinterpretation of seismic data, repeated seismic works, additional or repeated investigations in core sampling, repeat logging, additional drilling and other). One must admit that modern technologies of seismic exploration are more advanced than technologies used 10-15 years ago. What is the difference? The difference in that you will provide more than seismic data reprocessing only. What is more important - that one don’t have to believe to get a lot of new information after reprocessing. It is most probably that volume of additional information will be very low. Is all really bad? Of course no, if you plan to use these previous (old) results to project new works. It allows to reduce cost and time. 


The general purpose of the exploration seismic is to find out the perspective structures for pool boundaries substantiating and to drill successful wells (exploration or production). Some tasks (forecasting porosity, saturation and fluid contacts) can be solved with a help of inversion, AVO-analysis methods but the quality of forecasting strongly depends on the input data quality, the methodology of field works and the quality of processing and interpretation. The seismic exploration work has the specific depth interval of investigation and it is not so wide enough to investigate whole section of a field. Much money was spent without getting a result because that issue was not taken into account. This is of great importance for shallow depth layers (up to 1000 m). Study of these layers requires developing a special field project and processing. Resolution is another restriction for seismic data to use next. Resolution we mean that there is no possibility to identify layers with a thickness lower than some limit (it is vertical resolution), also we cannot identify some structures (!) if they have horizontal sizes lower than some limit (it is horizontal resolution). Our experience shows that layer should have at least 15-20m of thickness to be sure of forecasting (it is not a one-off  work with grants in seismic exploration area - Schlumberger, CGG, Fugro, Halliburton). By the way, a lot of oil&gas saturated formations have thickness lower than that value in Russia.


Why we need these examples? Of course, these examples doesn’t reduce or  increase the meaning of these studies. With a help of these examples we would  like to show you that there are no universal methods for the field study and we do not have universal stages of project realization (step by step for every field). But what is there? The common approach is - to try to establish the relation between the value of any study and what we are expecting to get as a result. Otherwise: the scenarios of the project realization we write by ourselves and provide this step at start of a project road. And after that we develop a program and studies to prove or to disprove these scenarios. But never other way around.


Geological model


So, we studied all the types of input data and methods of reserves estimation for oil and gas projects. What is the next step? We get logging, seismic and petrophysical data and use these one as an input data for 3D geological model. This model should form our view about the field.




3D modeling was going to be a part of our daily life and we cannot imagine the situation (new apartment remodeling, serious diseases intelligence, creating new automobiles and factories) when any global process is not connected with a 3D modeling. We should only bear in mind that dimensions of a model is of great importance only for fields which were being studied enough (more than 5 exploration wells with well tests, logs, core samples, 2D seismic with density not lower than 0.5 km/km2 or 3D seismic). In other cases building up the 3D model will not add an information and precision to the understanding of the object in comparison with the 1D/2D analysis of wells data but the cost and the time of the project work will be unreasonably increased.


Geological model is not a static object and is being changed during extension of our knowledge about the project. One can say more: at every step of the project geological modeling to solve the particular task – to estimate the reserves dispersion, to define typical geological zones for the development strategy, to putt well patterns, to plan well completion. Methodology and scope of work for every task is different. For example, for the exploration drilling we do not have to build up detailed geological model with dimensions of cells in XY lower than 500*500m. By the other side for producing well patterns on a pilot area the size of cells will not be more than 200*200m because in order to model effect of  non-uniform fluid influx into horizontal section of well with fractures requires setting specific parameters of model (e.g. transmissibility). Taking into account length of a horizontal section 600 m (for ex.) and essential non-uniformity of layer we have to use a model with cells size that not more than 60*60 m. 



The selecting system for well patterns (development system)


Where to begin if geologic data is not enough (often we have only the state balance sheet)?


The first postulate of correct approach is to say it does not matter what the category of reserves we have but we should estimate 2 scenarios at least: 


-        Conservative scenario (“gold” formula is – 100%С1+50%С2, however,  choice of estimate coefficient should be proved individually);


-        Optimistic scenario (add resources to upper formula with certainty factor 10-30%).


As a result we have 2 estimations. Next you can build distribution function (normal or log-normal, the result depends on your confidence in achievement of the edge values) and to calculate most probable value of reserves (P50). Described approach have rudiments from probable approach only. We call it minimally required. The minimally required does not mean that this methodology has essential restrictions. It works on the step when you are limited by a time and insufficient volume of required studies. On next step one will go on the full analysis of uncertainties when every value in formula of reserves estimation are shown as range of values (area, porosity, thickness, saturation coefficient, conversion factor, density and overpressure coefficient, temperature coefficient for gas reserves estimation).  




Reserves distribution histogram with estimations P10, P50, P90.


Methodologies of uncertainty estimation are described in a special literature. Many huge companies have a specialized software for uncertainties estimation and they train the staff. As for us, we have another issue – how to estimate the risks if you do not have enough information, budget and time.


So, now we try to see a case when the time and the volume of input data are enough to provide full analysis. There is 2D and 3D analysis in a global practice. The first one means estimation of uncertainties for every parameter included in reserves estimation formula and production profiles. The result is histograms with dispersion of every parameter. This analysis is reasonable for Pre-Appraise and Apprise stage. To implement this analysis we should have information from exploration wells and the some knowledge about field contours. Parameters for reserves calculation (ranges of parameters) is taken from the exploration drilling and regional seismic data. The common practice showed that common uncertainty is area of deposit. This characteristic is most significant and critically influences on a resulting reserves distribution. 


The undeniable advantage of 2D analysis is a minimum of input information and minimum time required for calculations. Need to mention sometimes this tool is not used right. We say about the choice of distribution function. Usually normal distribution (Gauss) with equal probability for values above and lower is the best choice than the average value. Though we often use log-normal distribution. Using this means that you can most probably get a value lower than the average. Thereby, log-normal distribution is a tool to manage risks and not to overestimate investors’ expectations.


 The most powerful tool for risks analysis is 3D modeling or 3D analysis (but this definition is not so preferable). This one is usable if you have high level and amount of input data.  By this analysis, one can identify with confidence the probable range of reserves and a consequence of production profiles. The advantage of this methodology bases on a huge amount of calculations (at least 50-100). By the other side, we can get the most probable scenario (P50). This scenario is a base for next technological and economical calculations. Nevertheless, we should bear in mind that this level is not a realization for some defined input parameters, this is a statistical estimation for next calculations inside TEOC.


The first postulate of right approach is to say that all uncertainties we can see in 3D models are going from uncertainties of input parameters (porosity, area, saturation, permeability, net pay). Keep in mind that increase of number of calculations will not result to changing the statistical consistent patterns. Why we should do 3D analysis? The first – we are living in a 3D world and we got accustomed to see 3 dimensions and time. So, 3D analysis allows us to choose right way to estimate uncertainties and risks – where to drill, what layer we should start a production and what sequence we should use. The 2D analysis does not give us the same possibility.


Another essential result is a possibility to estimate influence of varied parameters on result - reserves and production. This dependence is showed on a graph named as «Tornado-plot». This simple graph has a great value for this work because it helps to develop right project strategy (steps «Appraisal-Selection-Definition»). This graph also shows us that we should not investigate in all at one time but we should do it sequentially.



Sensitivity analysis


Production profiles 


The estimation of production profiles is not so simple work because uncertainties of reserves sum with selection of optimal strategy and field launching sequence.


Take a look on chart with different production levels that shown below. The dispersion is explained by different development strategies and not by a different volume of reserves. So wide dispersion is explained not only by chosen development steps but the variation of drilling schedule, estimation of wells production potential, wells’ design, completion system, selection of FPM agent and starting of it injection.


There is a big dispersion, isn’t there? Yes, but it’s ok…for some stages. This is a typical set of profiles for «Appraisal» and «Selection» stages of major project. This range should be reduced on next steps if we make correct project roadmap. The possible range of uncertainties we have to know at every stage of the project!



Uncertainty in production profile


So, the estimation of production profiles is a complex and primary task for oil and gas upstream projects. If we use systematic approach to estimate a project we have to:


1. To determine a scale of the project – probable levels of production profiles.


It means that before calculations we must understand what is the minima we could get. It is important for investors because the scale define tools for fund raising. Banks don’t like to fund money for small projects even if these give high key performance indicators.


2. Have the clear conception of optimal system and sequence of development, including the way of realization for pilot works.


The project team from start of project should “tune” investor for some big costs or essential investments before the production start. But the main question is how it should be proved? It should be proved from the point of view next possible losses. So pilot works not increase profitability of project in future but these works save an investor from big losses before main investments in field facilities.


Where we should do pilot works? We not discover America if we say that every field have some “sweet spots”, these spots are defined with a help of exploration drilling or some remote measurements like a seismic. We mean wells with the high rate. These particular areas will define effectiveness of your project in future. And these areas must be primary viewed on steps «Appraisal» and «Selection». If these areas have profitable economics then you can go on next steps of project. Other less known areas of field (or less productive) you can explore during production time when you will have profitable sale of oil/gas on main areas.


3. Definition of wells design, completion system, cost of wells.




Wells design and wells completion system with complex filters


The main attention should be aimed to typical wells design definition for every geological zone.


For example, wells with the simple filter construction (meshy, slotted or even open hole) is applicable for the solid and good cemented reservoir and allow not to choke stream in the well. But for weak-cemented formations (Cenomanian formations in West Siberia) where sand is one of “fluids” the essential task is to select the system which allow to hold sand before wellbore or to make some conditions in wellbore to prevent staying the sand on well bottom. In O&G world there are “smart well” and “smart completion” also. These tempting technologies are for lovers of high technology projects.  Our opinion is these technologies are not working good always and anywhere.  What is the criteria of selection? The criteria is rate and its dynamic in time.


4. Estimation of drilling schedule.


What is the main mistakes? It is clear that maximum production in first years is a good main tool to increase profitability of the project. Often a project team has been giving non-realistic numbers for regions with low degree of facilities. For example, using 4-5 rigs at the first year. Sometimes it is possible to make but what you will do if 2 rigs (for ex.) will get a result completely differed from the plan (lower rates or more worse – water-saturated collector or dry wells)? Mobilization cost, down time and demobilization are good tool to decrease project expenses. What is the optimal number of rigs to maximize efficiency? We have no answer, but we know - to find out the answer we have to start from the geology.


5. Definition of the development concept for surface facilities (depends on production profiles) including schedule and costs of capex.


Our experience shows that work requires at least 3 concepts. Primary questions is how many pipelines we need, do we need to build a pipeline or to evacuate by oiler, when to build CPF, to use pilot drilling or to enter wells sequentially, when to start injection? Each from these questions born different plan of construction and arrangement. There is one simple rule – one case should be the most simple (and probably cheap), the second one should be the most modern and highly technological (the most expensive usually). What is the base case? Base case will lie between the above-described cases depending on NPV, IRR or other economical parameter.


6. Estimation of economics for whole project.


This action point have to be done on every step of early described substantiating. Every calculation and optimization should be ended with calculations of project economics. You may say – “not all the things can be translated on money language. For example, how to answer on question: Should we do 3D seismic or drill one exploration well with core sampling or to drill one test production well with some another completion or trajectory”? We can answer – “It is possible to do through the volume of additional production we could get after this optimization”. Next steps are simple – we plan money for the pilot works, for example, to reduce drilling cost or increase the initial rate of the well or decrease the decline rate – there are many ways.


Estimation of production profiles is a multifactorial task and it requires taking into account many restrictions. The main restriction is economics - you have to keep in mind through the project. 


In real work the project team make dozens of iterations before the final production levels is being represented. Each iteration is aimed to optimize technical and economics parameters for every direction and project at whole.


Much as we would like but the selection of priority for substantiation of key development parameters (wells spacing, type of wells, allocation on section, wellbore length, action method) is not so simple for estimation of production profile as we would like.


For this reason now the most effective tool is an analysis of uncertainties (an equilibrium variation of influenced parameters).


Now for estimation of uncertainties exists many software solutions: Enable of Roxar company, MEPO of Schlumberger company, DMS of Halliburton company and some other. Selection of specific software depends on preferences of user and financial capacities. 




Results of estimation a production uncertainties with help of 3D models


These software solutions allow to compute automatically many cases of model with wide variation of input parameters. Using this we can estimate reasonableness of geological model with the help of history matching tool and estimating possible variation of reserves.


Summary – a task of forecasting production profiles is one from very complex tasks in a project. The approach should be based on 2 postulates at least: the first one is to minimize financial losses for investor and the second one is to develop tools to mitigate possible risks.




The drilling issues and completion issues is an important part and for some projects they are key issues. 


On the step «Pre-Appraisal» is enough simple estimation of wells quantity and wells design. But at steps «Appraisal» and «Selection» the detailed case study of wells design includes the next points:


-        Drilling and design of top part of wells including a solution of problems with lagging of permafrost and definition of minimal distance between wellheads (it influences on area of wells pad and cost of wells pad);


This is a primary restriction for drilling shallow wells with big horizontal displacements also.


-        Drilling and design of bottom part of wells taking into account a lift method and particularity of geology for deeper section (mud-loss, pipe becoming stuck, abnormally high pore pressure and other potentially complicating factors);


-        The maximal horizontal displacement and maximal length of horizontal section (it influence on number of well pads; a solution of this task require integral assessment including the type of rig, dependence of cost from wells design – horizontal displacement, length of horizontal section and integral cost of wells drilling and the building of well pads); 


-        The selection of wells completion system for production wells: how should we run and seat tubing string and pump? Any expensive well completion systems can be buried if you didn’t estimate primary risks and lessons learned at the beginning;


-        The configuration of horizontal section (different configurations may be viewed depending on geology specifics of productive formation: pure horizontal, uprise or topdown profiles).


A summarized version shall comply with the requirements of maximum economic attractiveness and minimum technological risks. The role of project team is being finding an optimal solution for these oppositely directed objectives.


A drilling schedule and design of wells allow to estimate cost of drilling and number of rigs. It is essential to define region of drilling, depth of drilling and pressure in rock formation correctly. The cost of mobilization/demobilization of rig may differ in many times for remote areas.


The drilling project activities is not possible on steps «Appraisal» and «Selection» without cooperation with potential subcontractors working in this area.