BEI ENERGETIQUE ET PROCEDE
STUDY OF SAGD PROCESS

 
Achievements II

 

This first simulation provided us some useful tips to get down to a proper and more realistic reservoir. Here several examples of configurations which can be encountered in a real reservoir:

    The first example is a modification in the reservoir porosity.

 


Simulating the same reservoir conditions, exempt a modification in the porosity showed us totally different results. And confirmed the need to work on a as high as possible porous reservoir


Same simulation, with two different porosity (1000 mD for the graphic top reservoir, 100 mD for the bottom one). The steam chamber as spread more easily in the first reservoir, heating more oil the in the second case.

    The second example provides a water cap at the top of the reservoir, preventing the steam to spread as it should

 

We understood that the presence of water (or gas, in a gas cap) can be negative for the recovery process, because this cap catches the heat and increase it own temperature, instead of warming the rest of the reservoir. (cf graphic aside) [3]




The presence of a water cap at the top of the reservoir hampers the steam chamber, which can not warm the rest of the reservoir

    We finally worked on two injectors/producers wells, and on a long time simulation, to see the steam chambers spill and meet.

 

It is also possible to see the residual and never recoverable oil, which lies in the bottom of the reservoirs. This critical amount has to be foreseen, to decide where to drill the wells. With this end in view, it is more than important to compare the price of an additional couple of wells (around 1 million $) and the amount of unrecoverable oil. [4]


Evolution of the steam chamber, after simulating 5 and 10 years

 


Residual unrecoverable oil in the bottom of the reservoir (after 10 years), with a 100 m width between the wells


Oil & Gas production forecast, for a 10 years simulation.


We clearly see the peak at the beginning, meaning that, after a short period of heating while shutting the wells, they became open, and the production began. As well, the decline of oil produced is evident, which means the field will provide less and less.

 

Until now, we have learnt the reservoir simulation, as well as the handle of our software. We also underlined some basic cases which need to be sought (eg. good porosity) or, on the contrary, shunned (eg. water or gas cap).

We also get use to a professional software, which we will certainly manipulate during our imminent reservoir engineering position. It was indeed very interesting to face realistic problems, and to search for reasonable & feasible solutions.


Heterogeneous reservoirs

The office plurality of all our results on simple, homogeneous and isotropic cases, enabled us to include/understand the mechanisms governing the evolution of the vapor room through the reservoir.
In the second time, and before being concerned with rates of recovery by comparing various works grouting, we were interested in the evolution of the fields of temperature on a heterogeneous case 2D.

The following simulation has enabled us to rub  with a more realistic case, where the properties of porosity, permeability, saturation of oil & gas… are anisotropic.

The first run highlighted a zone saturated with oil practically ever reached by the vapor room. Consequently, the production of this tank was relatively weak. That is explained by the fact why the producing layers were separated by an impermeable zone, which locally prevented the diffusion of heat in the reservoir





Porosity and oil saturation of the reservoir, in our heterogeneous and anisotropic pilot case.

 

The heterogeneity of this tank will generate zones where oil will be trapped, and where the vapor, blocked by layers of clay with null porosity, will not be able to be spread in the reservoir, and to heat oil like the ideal cases preceding.

Simulations realized with only one couple of injecting/producing well showed the impossibility to recover most of the oil, immobilized in the superior layers of the reservoir.

On the graph below, one sees clearly that the roadbase is not reached by the vapor room, and thus there is no production in this zone.




For stage with this problem, we placed a second nozzle steam well on the level of the not mobilized layer, to heat this part of the tank and to try to recover part of this oil. It is quite obvious that the drilling of an additional well required an important investment, which amounts to M$, but the volume of oil mobilized in this case makes it possible without any doubt to make profitable this expenditure.

 

Oil saturation in the tank after 5 years of production


This solution indeed makes it possible “to push” inaccessible oil in the first case, and to make it flow to the bottom of the reservoir.
One can observe these results besides on the graph this lower part (the production is more than doubled by 2010)



10 years forecasted production with the two different configurations

 

Comparison of production per solvent injection

Initially, we were interested exclusively in the impacts generated by the quantity of solvent in the reservoir. Indeed, it is possible to observe notorious variations in term of production, of injection but also of total economy (by taking into account the whole costs of the process).




Comparison of the production for a vapor injection alone, or a methane addition in variable proportion


Comparison of the profiles of production over 10 years for various mixtures vapor/methane

The preceding curves make it possible to obtain some invaluable clues as for the economic impact of the process of recovery.
It is noticed that the vapor injection alone, over one 5 years period, allows us to produce approximately 11 ' 000 m3 of oil.

The fact of introducing methane instead of the vapor, injected to 250°C, 25 bar, leads more or less to the same performances, that is to say same extracted volume of oil, but by introducing less vapor, therefore by saving energy, making the process all the more sparing. (Preceding Figures)

By making a fast economic study, in our example of heavy oil fields, we realized that the fact of heating 40 m3 of water with 250° requires approximately 1000 m3 of gas per day.
If one is satisfied, by agreeing a deficit of production limited, to inject only 75% of the volume of preceding vapor, we see that the impact in term of volume of gas is largely decreased (250 m3 per day in less, knowing that we will have to inject into the reservoir approximately 10 m3.

The profit is thus important, not only in term of cost (purchase and transport of gas), but also in term of polluting emissions (reduced in this case by 25% for same produced volume)

 

Comparison of production for various types of injected solvents

Now that we observed the impact of the addition of a solvent on the oil production, let's focus on the solvent itself. Indeed, the various hydrocarbons which one can choose to introduce into the tank have different properties of miscibility, and will make it possible, according to their chemical composition, to reduce viscosity more, and by voice of consequence, to recover more oil.

Our study was changed on a cocktail made up with 75% vapor and 25% of a light hydrocarbon (being given the results obtained in the preceding part). We thus compared the forecasts of production of our pilot reservoir, over 5 years, while injecting in turn:
            - 100% vapor
            - 75% vapor, 25% methane
            - 75% vapor, 25% butane/pentane
            - 75% vapor, 25% hexane/octane

           



Comparison of the prediction of production over 5 years, for various mixtures injected

The curves above clearly present a significant growth of the production for the vapor - hexane/octane mixture, which can be explained by the fact that the hexane has its temperature of vaporization close to the one we introduce the vapor into the reservoir, (250°C, 25 bars). Moreover, their close characteristics will allow a condensation of the mixture in periphery of the vapor chamber, and thus will facilitate the dissolution of hexane in oil.




It is thus noticed that the forecasts of production are twice more important when hexane is used, rather than methane or even of butane. The profitability of the investment will thus be largely improved; provided that the supply hexane is close (the cost of construction of a pipe between the refinery and the fields of production is often the number 1 problem)