☑ Data Scan
1) What kind of data can I check with Data Scan?
Data Scan performs checking of production data, as well as its consistency with well events and inflow test data.
2) What kind of data issues could be identified with Data Scan?
There are 70 validators divided into 4 groups that check data quality (unreported rates, ratios, inconsistent data, suspecious rates and ratios, constant rates and ratios), trend behaviour (non-monotonous cumulative production, sudden trend changes), production allocation vs. well events consistency (production without recorded perforation) and possible cross-flow between strings in dual string wells. This checking could be performed for wells, layers and field data.
3) I ran Data Scan and I got Data Scan Index value equal to 0.6 – is it good or bad?
Data Scan Index (DSI) value ranges from 0 to 1. The higher the index, the better the data quality is. DSI equal to 1 would mean data have no issues at all. But validators configurations and weightage also impact on the DSI value, so these factors have to be taken into account when comparing well DSI values.
4) What is DSI map?
DSI map shows total data quality indexes for all wells and interpolates values in between. This could help to visualize areas with poor or good data quality.
5) How can I correct my data?
After all issues have been identified, user may correct the data. It is possible to correct data manually, but for the convenience there is also DSI Toolbox available with some pre-set functions – average, minimum, maximum, inverse sign. Few trends could be applied as well – linear, regression. In addition, a calculator is embedded to the Data Scan which allows to type in any formula and apply it to selected data.
6) Are these corrections recorded somewhere?
Yes, user can track all changes made in Data Scan module by clicking “Tracking” tab in the module. All information about date, user made corrections, issue, property corrected, old value and new value will contain in the table under the tab.
☑ Production Allocation
1) What input data are needed for Production Allocation?
Main inputs for Deterministic Production Allocation are well production data, well events (perforation, zone changes), static zone properties at the well (permeability and net pay), reservoir pressure and well flowing pressure (if KHP/KHdP model is chosen). For Advanced Allocation some additional data are required: SCAL, PVT data and average reservoir pressure. Well flowing pressure could enhance the study but is not compulsory.
2) What is the difference between deterministic and advanced allocation models?
Deterministic allocation is production allocation performed with static or semi-static properties driven split-ratio (KH, KHdP etc). It provides same allocation factor for all phases and well has only one production allocation solution. Advanced Allocation is based on a multiple solution search engine. It takes into account watercut/gas-oil ratio evolution in time therefore resulting in multiple allocation solutions for each well. Both two-phase and three-phase models are available.
3) How many solutions can I get in Advanced Allocation?
Number of solutions is a user defined parameter, so it mostly depends on a time frame available for the study.
4) If I have PLT measurements can I use them in Advanced Allocation?
Yes, measured rates for a zone could be used as a PLT point which will constraint the search and reduce the allocation uncertainty. Also, if commingled well produces from a single reservoir for some period of time, it will become a constraining factor as well.
5) I have run Advanced Allocation with 100 solutions per well. How to choose the right one?
POSEIDON Production Allocation module contains a sub-module named Assessment, which aims to evaluate obtained well solutions and select one or few most possible. With contact tracking and recovery factor variance all solutions could be filtered and scored, which significantly facilitates the decision making process.
6) Can I run different Production Allocation models for different wells?
Yes, user may choose the allocation model to be run for each well.
☑ Remaining Oil Compliant Mapping (ROCM)
1) What format should the input data be in?
POSEIDON supports the import of .csv (comma delimited) files for production data, reservoir and fluid properties, Petrel XYZ format for static property maps, .las files for well log import, .dev files for well trajectory. Production data import from OFM is also supported.
2) ROCM requires production data, but what if our production and allocation data are not perfect? How much will it affect the result?
From our experience production data is never perfect and allocation always carries some uncertainty. On one hand, POSEIDON DataScan module performs the check of full production dataset, identifies missing/inconsistent/suspicious data and allows to automatically correct them with DataScan Toolbox. On the other hand, Advanced Allocation workflow allows to capture production data uncertainty. ROCM could have multiple realisations which helps to evaluate the impact of production allocation on the by-passed oil assessment and perform the sensitivity analysis.
3) Is a full static model required to do the ROCM study?
No, ROCM requires only a minimum set of maps: net volume height, pore volume height, HCPV height oil/gas, permeability, gross thickness, TVDss. Thus, it could be valuable to have a static model, but not compulsory.
4) Is a dynamic simulation model required for ROCM?
No, the dynamic model is not required.
5) If we only have 2D maps, how much would the ROCM study quality drop?
ROCM is a 2D process so a 3D model is upscaled to a 2D map on a flow unit basis. The vertical heterogeneity at the well is captured by the fractional flow functions which are conditioned by the permeability profile of each well. Therefore 2D maps are perfectly acceptable as part of the ROCM process.
6)How much does the result depend on the geological data input (maps)?
This is a very pertinent question. The advantage of ROCM is that it primarily utilises the trends in the dataset, and absolute values are less important than the relative contrast within the reservoir.
7)What if our 2D map is incorrect? Will it be misleading?
Geological uncertainties regarding the distribution of reservoir property trends should be tackled via the multi-realisation process. Different geological maps should be provided if there is considerable uncertainty regarding the reservoir quality distribution between wells (porosity, permeability, net thickness and saturation).
If the geology represented within the 2D map is inconsistent with the flow behaviour and /or the material balance calculations, then it will not be possible to obtain a good ROCM match outcome. This means that the 2D map should be revisited and the scenario provided is not a suitable representation of the initial hydrocarbon distribution. The ROCM process provides an internal consistency check to the geological modelling.
8) Could ROCM be applied for multi-zone reservoirs? Are there any projects confirming this experience?
Most projects we have worked on are multi-zone reservoirs (up to 100 layers in a field in Malaysia). ROCM is calculated on a flow-unit basis therefore production allocation must be performed as part of Advanced Allocation workflow (developed jointly with PETRONAS Research). It is also possible to use the existing zonal allocation from the field. Each zone is treated as an independent flow unit although cross-flow between zones or blocks could be modeled as well.
9) What if fluid and rock properties are not available or uncertain? How sensitive the ROCM results are to these data?
Uncertainty on fluid and rock-fluid properties is a common issue for most fields. For some fields the impact would be more, for others much less. If no SCAL data is available then some assumptions could be made and this uncertainty may be captured in multiple ROCM realisations as a sensitivity analysis. This would have to be done the same way for any analytical and/or numerical studies.