Print contacts The goal of Teics is to generate recommendations of the exact well interventions, which will lead to the reduction of expenses for development, growth/stabilization of oil production and stabilization of/decrease in water cut at brownfields. In separate cases we do not only propose some well intervention technique, but we also carry out a quantitative assessment, e.g. we specify the volume of a chemical to be used. You can request a full list of our recommendations.
Contacts
-
1
Send us a message 1@teics.com or call us +7 (347) 224-25-77
-
2
We will organize data transmission and we
will sign a confidentiality agreement.
-
3
Implement measures recommended by
Teics and achieve additional oil production!
Address: Moscow, Skolkovo Innovation Center, Nobel Str. 7, Office 75
Postal address: Ufa, Mendeleev Str. 145, 50
Oil
production
production
Well interventionis needed
Water
production
production
Well intervention solutions search
0%
Robotic well intervention solutions search
Robot scans the oilfields...
Well intervention planning is finished.
0%
0%
Additional
oil produced
oil produced
LOGICS
Output data
-
Withdrawals
and injections
per wellsRecommendations
of optimal
operation
conditions -
New drilling points
Inclined wells,
horizontal wells,
side wells,
horizontal
sidetracks -
Bottom-hole treatment
selective acidizing,
large-volume selective
acidizing, mud-acid
treatment -
Conformance control
including
volumes of
injection -
Hydraulic
fracturing
INSIM+ML
DATA-DRIVEN MODELS
AND MACHINE LEARNING
-
INSIM+ML
Solving tasks of
clusterization
and regression -
Engineering
logicsSteps of
computations -
ML
Physical model based
on the material
balance and
Buckley-Leverett
theory
Judging from today, we know what measurements were verified and approved in an incorrect way in the past. Using machine learning for clusterization of input data and their preparation, we re-approve the values of parameters automatically and reduce error in all subsequent calculations significantly. INSIM+ML physics-based model is INSIM-FT model improved with machine learning algorithms. At the output of INSIM+ML, data are sent to the input of the following mathematical learning models for the selection of well intervention techniques. All our experiments are already in the past. Now we return extremely high results following methodology created by us.
Input
data
-
Daily
data –
results gridDuring the whole
history – once,
during the pilot
project – on a
weekly basis -
Routes and
coordinatesDuring the whole
history – once,
during the pilot
project based
on drilling
facts -
Monthly
data,
operational
reportDuring the whole
history – once,
during the pilot
project – on a
monthly basis -
Well logging,
dynamic well
testingDuring the whole
history – once,
during the pilot
project – on a
monthly basis -
Implemented
repairs and well intervention
techniquesDuring the whole
history – once,
during the pilot
project – on a
weekly basis
Teics tries to use historical data of field development as much as possible. However, due to various reasons, we do not always have information in its full scope. That is why our requirements are more of suggestions. We are ready to discuss a possibility of working with the information that is available. Having requested a list of input data in advance and having prepared, you will speed up the start of our joint work significantly and get practical results much faster.
RESERVOIR PHYSICS AND MACHINE LEARNING
Having replaced the most part of manual work, Teics got an opportunity
.to actively work with Big Data and apply machine learning (ML) algorithms in its work
For robotization, the Teics team used not only mathematical tools, but also improved
data-driven models (INSIM+ML) in order to go from the “black box” notion
.and add physical sense to its solutions during development control
Retrospective analysis
In this option of pilot tests, an extracting company gives Teics data on the history of development until a certain moment in the past. As a rule, this is half a year before the present day. Then the extracting company provides a list of measures taken during the last half a year, without indicating their effectiveness. Teics forecasts the productivity of the measures taken based on the historical moment in the past. Comparing the Teics forecast with actual effectiveness, we get the numbers indicating the practicality of introducing "Teics One" in the company. Besides measures taken, the retrospective analysis can also include measures proposed by "Teics One" for the selected historical moment. As a result, such kind of pilot tests enables to see the prospects of introducing "Teics One" in fully safe mode of research in a prompt manner (about half a year).
Teics measures
In this option of pilot tests, Teics together with an extracting company prepares a monthly program of measures to be taken for wells, and the result is an additional amount of oil extracted for the selected period (as a rule, from half a year to one year). Data on the history of development are handed over up to the present date and are provided on a weekly basis. A program of measures is prepared once a month. From the present date, based on the dynamics of the previous years, a forecast for oil extraction is formed one year ahead, and the actual data are compared with it, which defines the effectiveness of robotic development. Such pilot tests include a theoretic probability of mistakes when selecting such measures. As our experience shows for already implemented pilot projects, Teics solutions turn out to be more accurate than the existing methodologies, however, joint discussion with extracting companies on a monthly basis related to measures selected from both sides and their approval leads to growing trust and high transparency in the relations of the partners. Such kind of pilot tests enables to obtain additionally extracted amounts of oil in one year - one year and a half (it depends on the period of measures taken).