3 Fracture Vuggy Carb Res

73rd EAGE Conference & Exhibition incorporating SPE EUROPEC 2011 Vienna, Austria, 23-26 May 2011

P116Fractured-vuggy Reservoir Characterization ofCarbonate, Tarim Basin, Northwest ChinaP. Yang* (China University of Petroleum / BGP-CNPC), Y.L. Liu (BGP-CNPC),H.Y. Li (BGP-CNPC), G.J. Dan (BGP-CNPC), H.T. An (BGP-CNPC) & Y.M. Shao(BGP-CNPC)

SUMMARYFractured-vuggy carbonate reservoirs in Tarim basin have special reflections on seismic section termedstrong beadlike-reflections, which are easy to be located and are important drilling targets at present. But itis very difficult to distinguish the high yield ones from the common ones because of their complicatedinner structures and fluid conditions. This paper tries to use the term of fracture-cave body to describe andclassify this kind of carbonate reservoir. Enlightened by modern karstology phenomena, it begins reservoircharacterization with studying reservoir distribution pattern via reservoir-development control factorsanalysis, and points out that the fracture prediction, volume calculation and hydrocarbon identification areessential to locate profitable targets. Both coherence analysis and azimuth anisotropy analysis are used topredict the fractures. Based on the iteration of acoustic inversion and forward modeling, it gets therelatively proper parameters for volume calculation. Both poststack and prestack techniques are combinedwith geological analyses to predict the oil/gas bearing conditions of the targets. The raising drilling successratio shows that the new work flow and the techniques used in it are effective.


Introduction

The fractured-vuggy carbonates are the main reservoirs whose main storage spaces are secondary pores and fractures (Pang, 2010) which make the reservoirs highly inhomogeneous. This kind of reservoir will present a very special reflection on seismic section termed strong beadlike-reflections (SBR). SBRs are easy to be located and are important drilling targets at present. But it is very difficult to distinguish the high yield SBRs from the common ones because their appearances are very similar. Many geological (Liu, 2008) and geophysical (Wu, 2009) techniques are used to do reservoir prediction and great progress are achieved in recent years (Zhou, 2009), but most of the current techniques treat the SBR as a single geological body and usually overlook the complicated inner structure of it. Our study shows that the inner structure of a fractured-vuggy reservoir will influence the reflection features of it, and the prediction accuracy will decrease when it is overlooked. We try to use the term of fracture-cave body (FCB) to describe and classify this kind of carbonate reservoir and find the relationship between geological factors and reflection features. After that, geological backgrounds which influence the geological features of the FCB are analyzed. To achieve a higher success ratio and production ability, volume calculation and hydrocarbon identification are necessary besides reservoir prediction. Some new techniques such as azimuth anisotropy analysis and AVO analysis are used in this process.

Geological Texture models

Tarim basin locates in the northwest part of China, and it is the largest Paleozoic marine facies craton basin in China with the total area of 560,000 km2. So carbonates E&P are very important in this basin. Carbonate reservoirs of Tarim basin have some significant features which distinguish them form other ones of the world. The Ordovician or Cambrian carbonates are very deeply buried (usually more than 5500m) which result in the matrix porosities of these rocks being very small (less than 2%). The fractured-vuggy carbonates are the main reservoirs whose main storage spaces are secondary pores and fractures. Because of the great acoustic-impedance difference between the reservoirs and the surrounding rocks, the fractured-vuggy reservoir will present a very special reflection on seismic section termed strong beadlike-reflections (SBR). Because of their high amplitudes and narrow extent, they are very obvious both on seismic sections and RMS amplitude maps (Figure 1).

Figure 1: RMS amplitude map (left) and seismic section (right) example of typical strong beadlike reflections (SBRs). The red-to-yellow points in the map are the plane projection of the SBRs, and the seismic line AA pass through 4 SBRs.

Although many SBRs had brought to us good reservoir, some of them did not bring to us expectative petroleum output. In fact, besides different amount of oil/gas, the SBRs can also bring to us mud, water, or just tight rock. That is because the inner structure and fluid conditions of the fractured-vuggy reservoirs to which the SBRs corresponding are very complicated, no need to say the connection conditions between the SBRs which decide whether the petroleum in them can be produced through a single well or not. Outcrop data, well data and seismic data are used to build up of the geological texture models of the fractured-vuggy reservoirs. We use the term of fracture-cave body (FCB) which is somewhat like the term of paleocave system. But FCB pay more emphasis on the integration of fractures and caves as


well as their common function of storage spaces. Simply, A FCB is composed of one cave or several caves as the main body and some fracture zones surround it or between them. It can be in any scale. There are 3 basic FCB models: they are honeycomb model, hamburger model and pineapple model (Figure 2), separately. The type of a FCB is ultimately decided by the dissolution and collapse degree of the fractured-vuggy reservoirs, and the three models mentioned above have a relationship of increasing development degree shown in figure 2. Generally, a SBR is corresponding to a seismic scale fracture-cave body (FCB) that can be range from 20m to 200m in vertical and 50m to 500m or even much longer in horizontal.

surrounding

rock

Fractured

collapsed zone cave

Dissolved

fractures

Pineapple modelHamburger modelHoneycomb model

Increasing dissolution and collapse

surrounding

rock

Fractured

collapsed zone cave

Dissolved

fractures

Pineapple modelHamburger modelHoneycomb model

Increasing dissolution and collapse

Figure 2: Schematic diagram showing 3 geological texture models of the fractured-vuggy reservoirs.

According to forward modeling, we know that a SBR is the integrated response of both fractures and caves inside a FCB, that means the inner structure as well as its caves and fractures of a SBR are unrecognizable. But there are still some ways to judge qualitatively whether a SBR is good enough to be drilled. Among them, the most important rule is: the amplitude of a SBR has a positively relationship with the total porosity of the FCB as well as the volume of it. It can even be deduced (not be proved) that the amplitude of a SBR has a positively relationship with the total pore space (the product of total porosity and volume) of the FCB. That means, in some conditions (for example, the volume difference is not too big), the seismic responses of a small FCB with higher porosity and a big FCB with lower porosity may be the same. So, it can be believed that the stronger the reflection of a SBR is, the better the reservoir of the corresponding FCB is. Also based on forward modeling analysis, we know that providing the total pore space were the same, a FCB of honeycomb model would have smallest amplitude in its SBR while a FCB of pineapple model have the biggest. It means that in different areas, where the types of FCBs may be different, we should use different amplitude threshold for reservoir prediction. If a cluster of FCBs are connected by open fractures, we call them a fracture-cave unit (FCU). In this condition, we can use one well to produce all the petroleum in different FCBs. Using the concept of FCB and FCU can easily interpret the unpredictable drilling results, such as only meeting a 1m cave while drilling a big SBR, a FCB yielding such a huge amount of oil that it is obviously beyond the volumes of it, and so on.

Reservoir development pattern

Since the type of FCB has influence on seismic reflection but we can not distinguish them through seismic, we must do geological analyses to solve the problems. According to modern karstology we know that the forming of FCBs in a carbonate layer has close relationship with tectonic stress after sedimentation, current fault system, paleo-landform, and especially, paleodrainage pattern. The longer the carbonate strata were exposed and weathered, the more likely the FCBs will belong to pineapple model. Many techniques such as 3D visualization and trend surface are used to characterize these geological features. In this step, we also know that whether a FCB is filled with mud or not depends on whether it is located under a surface channel or under a residual hill, and the shape of the residual hill may has important influence on the accumulation ability and production ability of the oil/gas pool.


Then the reservoir development pattern becomes clear (Figure 3). Generally, regional tectonic stress and the faults formed under the stress are the basis of forming karst landform and paleodrainage; paleo-landform and paleodrainage react on each other through weathering, erosion, transportation, and sedimentation; during this process, the carbonate reservoir forms along and around the channels and FCBs appear in the places where the reservoir developed best. Usually, a FCB under a residual hill will be unfilled, and if its SBR is strong and big enough, it can be selected as the potential target. But before well proposing, volume calculation and hydrocarbon identification must be executed to reduce the risk further.

Figure 3: Schematic diagram showing FCB developing mechanism.

Volume calculation and hydrocarbon identification

We calculate the volume to estimate whether a FCB is worthy drilling or not. Since the volume of a FCB cannot be calculated like what we do to an ordinary sand reservoir, we use an iteration of AI inversion and forward modeling bases on wave equation to get the proper parameters for FCB volume calculation. If the forward modeling profile is very similar to the original seismic section, we can believed that the AI inversion result is acceptable and the following parameters of velocity and porosity in geological model is relatively correct. The other important parameters include volume calibration factor, porosity calibration factor, and threshold of AI value. Among them, the volume calibration factor is got from a series of forward modeling and the other two parameters can be got from cross plot analyses of well data and seismic data. Using these parameters, following the steps of volume adjustment, porosity adjustment, time-to-depth conversion, and reservoir sculpture in 3DV, we can get the final volume of the FCB at last. Again, since different places have different FCB types, the above parameters should be vary with their geological conditions. Since the connection condition of different FCBs is important to oil/gas producing, we pay many attentions to fracture prediction to judge whether a cluster of FCBs are connected together by the fractures and form a FCU. The techniques employed in this step include qualitative methods such as enhanced coherence and curvature as well as quantitative method like anisotropy. Attention must be paid that not all the fractures in a FCU are really contribute to the connectivity of FCBs for some of them are not open. The oil/gas pool volume should be the total volume of all the FCBs in a FCU. To reduce the drilling risk, hydrocarbon identification (HI) is important before proposing well location. Both frequency anomalies in poststack domain and AVO anomalies in pre-stack domain are employed in HI. Because the reservoirs are deeply buried, the lithology is carbonate, and the strata are highly inhomogeneous, the seismic data quality is sometimes not good enough to ensure that the results are correct. In this condition, geological analyses must be combined with geophysical


prediction to get the reasonable results. For example, we must take geological position into consideration for its important influence on petroleum acumination. Figure 4 (a) shows that in the low position marked by the black ellipse, although there is good reservoir, but there may be no petroleum according to HI result. This means that oil/gas acumination is not controlled by the reservoir alone like what was thought before, but also controlled by the basic rule of gravity differentiation. Figure 4 (b) shows that a water well (W11) in lower position is saved by its sidetrack (W11C) in upper position. HI plays an important role in determining whether W11C should be drilled or not.

Figure 4: hydrocarbon identification plays important roles in avoiding risks or reducing loss. (a) Difference of reservoir prediction (left) and HI (right) in structural low. (b) W11 (water) and its sidetrack well W11C (oil & gas).

After all the above studies are finished, the well location can be easily pointed out. The key points to well location selection includes: a relatively higher position, stronger reflection, more fractures, bigger volume, and clearer hydrocarbon response. By this way, the well drilling success ratio raises from 40% of year 2008 to 76% of year 2009 and 2010.

Conclusions

Based on above analyses and application results, the following conclusions can be drawn: 1) Using the new concept of FCB and FUC to describe and classify the fractured-vuggy reservoir in Tarim basin is necessary and effective. 2) Because FCBs type has influence on seismic reflection and it is mainly decided by dissolution degree, it is important to study the paleo-landform and paleodrainage pattern besides carrying out common reservoir prediction to help improve the accuracy. 3) Although it is very difficult and totally different from clastic reservoir, calculating the volume and predicting the gas/oil bearing condition of a FCB is necessary and realizable. It helps us avoid the drilling risk, reduce the loss, and increase the success ratio effectively.

Acknowledgments

The author thanks Tarim Oilfield Company for its authorization to publish this work, and thanks Mr. Wang JZ, Lu D, Zhao JX, Zhang XZ and Zhang RJ for their help in preparing this paper.

References

Liu LF et al., [2008] Reservoir types and favorable oil-gas exploration zone prediction of the Upper Ordovician Lianglitage Formation in Tazhong No.1 fault belt of Tarim Basin. Journal of Palaeogeography, 10(3): 221-230.

Pang XQ et al., [2010] Main progress and problems in research on Ordovician hydrocarbon accumulation in the Tarim Basin. Petroleum Science, 7(2):147-163.

WU XS et al., [2009] Difficulty and countermeasures in carbonate paleokarst reservoir prediction. Journal of China University of Petroleum (Edition of Natural Science), 33(6): 16-21.

Zhou XY et al., [2009] Cases of Discovery and Exploration of Marine Fields in China (Part 12): Lunnan Ordovician Oil-Gas Field in Tarim Basin. Marine Origin Petroleum Geology, 14(4): 67-77.

Documents

3 Fracture Vuggy Carb Res