Gravel packing is one of the commonly used sand control techniques in long openhole horizontal completions. Although numerous wells have been gravel packed with either one of the two placement techniques (α/β or shunt tubes), achieving a complete pack with sand control screens that have inflow control devices (ICD) can be challenging to say the least. This is because of the high pressure drops needed for the fluid to flow through the ICD into the base-pipe/wash-pipe annulus at the typical rates used for gravel packing, resulting in significant pressure rise and exceeding the fracturing pressure of the formation. Although using a screen without an ICD at the bottommost joint would certainly allow the α-wave to reach the toe, the pressure buildup will occur as soon as the β-wave proceeds upstream of that joint, which will either result in formation fracturing and bridging or significant rate reduction and α-wave height increase (near the heel first), both events resulting in a premature screenout. Ocelote field in Colombia requires sand control, and gravel packing and standalone screens (SAS) have both been used in openhole horizontal completions in various parts of the field. An additional challenge in this field has been the premature water breakthrough from the aquifer leading to very high water cuts after only a few months of production. ICDs have successfully been implemented in SAS completions to address this challenge in other parts of the world. Similarly, gravel packing has been successful in controlling sand in parts of the field where SAS was considered unsatisfactory, although water cut in most of these wells increased to 70 to 90% after only 3 months of production. In this paper, we present a novel technique for wells that require gravel packing for sand control and ICD functionality for managing water encroachment. Detailed in the paper is the first application of the proposed technique in Ocelote field, which resulted in 100% pack efficiencies using the water packing technique as well as significant reduction in water cuts, with substantial improvement in the project economics.

Oil production from the Llanos foreland basin in Colombia is seriously affected by sand production and high water cut. Sand production has been successfully controlled with horizontal openhole gravel packs; however, early water break-through causing water cuts as high as 90% remained a challenge. As a result, operators were forced to shut down some new wells just a few months after placing them on production. A new production logging approach was performed in five horizontal Hocol S.A. (subsidiary of Ecopetrol) wells. In terms of reservoir behavior, the results revealed the following: rapid swelling swell packers were swelled using a new swelling fluid that allowed the swelling to take place before gravel packing the well. Once the swell packers were swelled, the gravel-pack operation began and it was confirmed that every segment was gravel pack, even after the swell packers were in place, thanks to the implementation of downhole gauges throughout the completion. Once the well was completed, the production group installed selective production and inflow control strings to produce the well.The first completion installation was performed in July 2013. The operator performed characterization tests for each productive sand, with results for each sand indicating successful zonal isolation. The results also provided valuable data for reservoir characterization and selection of reservoir management strategies. The addition of new reserves confirmed the value and effectiveness of this new completion technology and its application.

A reservoir management strategy is presented, based on a pressure-production behavior analysis, which was carried out using data from a Llanos Basin's field in Colombia, applying the concept of Darcy's Law and Babu-Odeh Equation for verticaldeviated and horizontal wells respectively. The analysis can be applied on dead oil reservoirs (P res > P b ) providing a surveillance tool for making right decisions regarding the drawdown application as the well is produced, and thus maximizing the recovery factor.The methodology allows the estimation of the dynamic evolution of skin factor, pressure drawdown and water saturation by well, which are related as well to variables such as water cut (W cut ) and P wf for an easy handling and understanding of the field for geoscientists, reservoir and production engineers.This methodology could be applied without the necessity to build a previous simulation model, but in this case a simulation model was carried out, in order to validate the input data, obtaining a good history match. As a result of this, the methodology and numerical simulation model are complementary themselves.For this case, it was possible to define the proper/wrong application of drawdown increments that improve/reduce (mainly because of sanding) the productivity of vertical, deviated and horizontal wells for specific ranges of water cut. 2 SPE 132698 Simulation Model The current simulaton model is based on the layer cake model because all the 16 drilled wells are currently along the fault only. The main characteristics are: Total cells: 52x318x15 = 248,040 Active cells: 46,514 Porosity range: From 13 to 30% Permeability range: From 100 to 4000 mD K y = K x K z = [1 to 15%] K x Initial Water Saturation: 22% Initial conditions for equilibrium Black oil (P b = 167 psi; R s = 20 scf/stb) 2 phases flow P i = 1640 psia @ Datum: -3418 ft TVDss WOC = -3452 ft TVDss P cow = 0 psia Aquifer Volume: Infinite Primary Control: Liquid rate Secondary Control: Historical values of Bottom hole flowing pressures (P wf )The history match was based on tuning of K z /K x ratio and the oil-water relative permeability curves (K ro, K rw ) implemented. The (K ro, K rw ) curves were taken from the simulation model of two analog fields located at the Llanos Basin too (where the company is partner). This set of curves was artificially created, but finally it works. We tried several sensitivity runs using Corey type curves without success, but it is pending to incorporate the curves from SCAL analysis. SPE 132698 Calculation of: φ, Ka, Sw and Vclay, per well Assumes Cutoff values of: φ, Ka, Sw and Vclay, per unit/sub-unit Perform Summation process using % φH and % KH and obtain: Average values per unit/sub-unit φ, Ka, Sw and Vclay. Calculation of the total average of absolute KH value per well. Conversion of the total average of absolute KH value per well to effective KH values per well using the Krow curve defined for your reservoir Effective KH from petrophysics = Efective KH from PBU ? No Yes Define Skin from Darcy's Law with initial ...

Hocol Colombia operates in Ocelote field at the eastern part of Colombia, a land called as Llanos Orientales (Eastern Plains). In this field Hocol produced from Carbonera formation observing problems with formation deconsolidation. Normally, gravel packs jobs are performed on vertical wells with sustained good results. However, new reservoir studies showed that horizontal wells with regular length extension (around 1000 ft) would be more profitable for this field. In order to get this objectives, Hocol design a horizontal wells Program to drill and test alternatives to get good completion of these wells. The first well was completed and gravel packed using conventional technologies and a 35% of pack efficiency was achieved. The objective for the second well was to get at least 75% of packing efficiency. In a synergetic team effort, service companies proposed operator company new technologies that are used in other places but never were applied in Colombia. After simulations and technical meetings, the Operator decided to test two new technologies to try to get their operative and technical objective. This paper presents the technical and operative fundamental used to support these changes and the results obtained with theses applications.