Eclipse step by step

[reservoir_engineer]

dear Temr, shakespear
this is very good thread, i will fllow this thread carefully.
i used Eclipse and petrel for 2 years uptill now.
i will learn more and more from disscusion, and i hope i can add points for our brothers.
for Equlnum, pvtnum, satnum
it is described perfectly by temr
but it is little bit comlicated keywords
for example EQULNUM
you have to determine how many equilibration regions you have using key word TABDIMS in RUNSPECS section
then you have to set the kodes 1,2,3,.. for all cells using key word EQULNUM in region section.
then you enter each region properties using Keywords EQUIL,.. in solution section

ok, let's make it easier in the attached PPT file
go ahead Temr
Regards,
___________________
reservoir_engineer

[zeliustitan]

good tutorial.. but, why discontinued?

[temr]

First of all Thank YOU all to all for you comment
To Shakespear : I will review in details schedule section and how schedule calculates Kh , CCF, Ro with examples in excel file but as i said before i want do this thread step by step but section schedule is last )). but if you have specific question i will be glad to answer them. Btw i speak Russian and heriot and watt is good recourse for knowledge
SEUNCAROLINE , brainmolester : glad to see your input but do not forget to ask questions if any .
reservoir_engineer- agree with you but the DATA file compile will be discussed later because the structure of this thread is first to give general overview of input data , some physics stand behind process in reservoir and then how to write all this in DATA file. The big problem of all manuals is they only teach how to press the buttons)). Again i will be glad from all you help and again thank you for you support.
zeliustitan - i will but it takes time to write and prepare sets

[temr]

So described in Chapter 1 and 2 data allow to simulator
1. Position in 3D cell with having 8 coordinates and depth of each corner
2. Calculate Area of Faces of cell for transmissibility's calculations.
3. Calculate NTG, Pore Volume, Initial water Saturation ( detail will be discussed in Initialization chapter).
4. Position the faults and define they permeability's.
This all except water saturation is static parameters . But simulator is used for dynamic purposes so then we need do describe the properties of changing parameters.
They are
1. Pressure of gird block
2. How much grid block rock can compress vs pressure
3. Saturation of each phase in grid block
4. Static Property of each phase like viscosity, formation volume factor, dissolved gas-oil ratio, bubble point versus pressure ( as eclipse 100 is isothermical we assume that temperature is constant)
5. Dynamic properties - how Fast one phase flow versus reference phase versus saturation of reference phase
1. To tell simulator pressure of grid block you need to supply table for each Equilibrium region (EQLNUM) giving at least to depth and according pressure ( bar for metric system, 1 bar = 0.986 Atm=14.504 Psi(field) ) like in example below.
400 70
500 90
This example shows that on depth 400 m pressure will be 70 and on depth 500 m my pressure will be 90 bar
Having two points Eclipse then can linearly extrapolate pressure for not given depth for example the pressure for depth 700 m will make 130 bar (pic1)



Exercise 1.
In excel build graph and get linear trend with that trend check pressure for 700 m.
2. Rock Compressibility
When i start to make my models one of my major question why do simulator needs rock compressibility. This question was open till i wrote my 1 material balance program . So i will try to explain it brief and simple.
For each reservoir there 4 drivers ( the forces that help oil to get to the surface) and all of them calculates by eclipse
1. Depletion drive- energy comes from expansion of volume of oil due to dissolve gas
2. Segregation Drive- energy comes from expansion of Cap drive
3. Water drive - energy comes from outcoming water ( aquifer or injected)
4. Expansion drive - energy comes from expansion of rock. to calculate this you need to supply rock compressibility

The dimension of rock compressibility is 1/bars or bar -1 and mean how much rock will change in volume if pressure drows to 1 bar.
So why should we pay attention to this small value ???
NOTE!!!!!!!!!!!!!! The Rock compressibility parameter is one of those who effects to Average reservoir pressure . And this is one of tuning parameter during history matching.


3. Saturation of each phase in grid block ( detail description in Initialization chapter)
First of all at any time Sw+So+Sg=1
Where Sw - water saturation
So- oil saturation
Sg - gas saturation
Before Going to initializations lets make brief introduction for behave of Fluid in pore media
This is very important to any reservoir engineer to understand physical process in the pore and forces which allow flow in pore media

One of most important forces is capillary pressure
Speaking in general capillary pressure (PC) is a force which is opposite to displacement force.
Capillary pressure is the function of pore radius, tension forces of phases, wetability of phases
The main thing you need to know about it is that for each type poro medea with increasing displacement force at the end the will be irreducible saturation of phase which you try to displace.
Swc- Saturation of water connate.
The question you might have in your head could be why there is no Soi( Saturation of oil connate) -
this is because, when oil start to penetrate to oil cap and as it lighter than water , the oil starts to displace water and in this case displacement force will be Archimedean force of oil and capillary pressure of oil which try to stop oil going up.
For best example lets do the following
Lets take cap and imagine that it is pore if you will pore oil(vegetable for ex) and then pore water you will have two-phase immiscible system. If you seal cap on top and reverse it you will have the model of pore during formation of oil cap the oil start to displace water . But due to big pore our capillary force will be very small and oil will displace water almost for 100%.
So how simulator calculates capillary pressure of water-oil phases
PC= (rw-ro)*0.098*H
where
PC is capillary pressure in bars
Rw- formation water density, g/cm3
Ro-formation oil density, g/cm3
0.098- gravitation constant
H- oil water column high , m
So after simulator now PC he can calculate Sw from the capillary pressure table you supply as part of input data how the example below shows pic 2

We supply to simulator lab data base on which simulator gets trend showed on Graf1
Now simulator can build Table 1 were Swc is 0.391
He calculates PC and after he can find required Sw from table 1
NOTE!!!!!!!!!!!! Changing Swc value( in our example 0.391) can impact on changing reserves in place . This can be used during matching reserves in sim model to geological model
After Simulator know Pressure of Block and Water Saturation he can calculate saturation of other Phases

 

4. Static Property of each phase like viscosity, formation volume factor, dissolved gas-oil ratio, bubble point versus pressure ( as eclipse 100 is isothermical we assume that temperature is constant)
To run 3 phase you need to tell simulator how each of your 3 phases behave in different pressure.
The exact data for our test run will be described in more details
in general you need to supply simulator density at standard conditions, formation volume factor , solved gas ratio , viscosity and other parameters to be able simulator to calculate following
Bubble point pressure
Define Free phases in cell
Define saturation of each free phase
Define volume of dissolve phase ( gas)
Viscosity of phases for calculating mobility of phase

5. Dynamic properties - how Fast one phase flow versus reference phase versus saturation of reference phase
To enable simulator Calculate multiple phases flow, you need to supply relative permeability's curves
In general it is lab data but you need to understand what it consist of
As far as i remember Special core analysis a made in following steps
1. Saturate core sample with oil
2. Inject water
3. You measure input volumes of injected water ( base on it you can calculate average saturation of water in core )
4. Measure output volume of oil and water ( base on it you can calculate fraction o relative permeability of each phase)
So the relative permeability table is table were you give saturation of injected phase and fraction of initial phase and injected phase pic3

In this graph
Sw - water saturation
Krw - Relative permeability's of water
Krow - Relative permeability's oil- water
Pcow - capillary pressure of oil-water phases
During special core analysis you also get other MOST important parameter is Sor - Saturation of oil residual.
This parameter means how many of "lost" oil is in pore . "Lost" means that you will never be able to recover it .
In my example almost 20% of pore volume is 'Lost" oil
NOTE!!!!!!!!!!!!!! Form this graph you can get Sweep efficiency factor
Sweep efficiency is maximum recovery factor you can get from this type of rock saturated with this type of oil and calculates by following formula
Sweep=(1-Swc-Swr)/(1-Swc)
In my example the maximum Rf i can get is 64%
From this data Eclipse can calculate fraction flow, total mobility, permeability ratio of each phases in each time step taking in to account that eclipse knows pressure and saturation.(pic4-6)
on pictures
kro-water permeability
Krw - water permeability
mw- water reservoir viscosity
mo oil reservoir viscosity

pic4


pic 5


pic 6

So next chapter will be brief DATA file section description and we will build the model without wells for now with data set (petrel project) available on my thread.
Chapter 4

link to pictures

[link Point to another website Only the registered members can access]

[temr]

Chapter 4
In this chapter i will briefly describe eclipse data file structure and build data set for the test runs.
First of all a little of history
Eclipse was invented by 3 professors and the bought by Schlumberger as Petrel
As it the time the Eclipse was written ( on fortran if i am not mistaken) years past the interface of eclipse and the data it used were remain on level when we played tetris on i486 machines but now we have petrel and i will show later how we can substitute Flowiz with petrel
So once you double click on eclipse link you will return to

On the top menu you have following
Eclipse - or eclipse black oil for 3 phase isothermal black oil runs
Eclipse 300 - or compositional for thermal compositional run the different with eclipse 100 that you not only apply temperature but your oil or gas present as componential mixture ( pseudo component like C4, C5 etc) . As the amount of phases increase from 3 to 16 or 20 the time runs for eclipse 300 are much more comparing to eclipse 100. Basically this option required for condensate field and reservoir treatment like steam or others .
Frontsim - is the simulator but comparing to eclipse it is as called front line simulator. If eclipse solves nonlinear equations in every cell of reservoir , Frontsim 1 calculate calculates pressure gradients from to well and only then calculates fluid property only in cells which are crossing that pressure line or In other words pressure connections to well. It is good in water flooding problem and big field problems. ( as i now this only theoretically, you are welcome to add if i miss something)
Flogrid - postprocessor for visualization. It is very week so we will not use it Instead we will use petrel to see results of simulation.
Office - tool that help you to automation creation of case , I do not use it and personally do not recommend it but again it is my personal opinion
SCAL - tool for making relative permeability curves and capillary pressure. I use excel for this purpose later i will show example of excel file
PVTi- tool for modeling pvt properties , did not try this use Petroleum expert or excel
Schedule - tool for creating well connection and production history , good tool but again excel is better
VFPi - tool for creating Vertical Flow profile curves. They are required to able eclipse to calculate Bottom hole pressure for tubing head pressure. Again use excel
Manual - this stuff u should read always
So for good runs all u need is Petrel, Eclipse , and Excel
Data file
NOTE!!!!!!!!! in some application " " empty line that you see in excel or word could not be actual empty line txt editor so be careful when copy and pasting data form excel to text file for this purposis use advance text editors like notepad++

[link Point to another website Only the registered members can access] As i told you before eclipse was invented a lot of years ago and approach for loading data was not change since it began commercial.
The approach for loading data is script or command text file which have extension .data
The syntactic of data file quite simple
1. You give Keyword
2. You supply data in a form of table
3. You terminate keyword and table by / (slash)
4. If you want to make comment use first -- then write comment
I this chapter i will link all keyword description to Eclipse reference manual which is available from eclipse or from directory C:\ecl\2009.1\manuals .

Data file structure


The script file shoul consist of several section
1. RUNSPEC
2.GRID
3.EDIT(OPTIONAL)
4.PROPS
5.REGIONS
6.SOLUTIONS
7.SUMMARY
8.SCHEDULE
All this levels are amust and each keyword has its own predefine section.
NOTE!!!!!!!!!!!!If you will use keyword in wrong section eclipse will give you error and probably stop the run
To see what keyword should be run in what section you should open eclipse reference manual find key word and see what section it is belong too pic1

In example on picture keyword DIMENS should be in Runspec section . if you will use it in Grid section eclipse will give you an error.
So lets describe in general what are minimum required information for make 3 phase run
1. RUNSPEC - in this section you need to supply following information
TITLE- case name
DIMENS - dimension of your grid in i j k directions( look to Chapter 1 how to get this from petrel)
METRIC - type of system ( metric of field)
GAS- phase of run
OIL-phase of run
WATER-phase of run
NOSIM - keywords telling eclipse not tu run simulation just read data in test mode for errors
So for 3 phase run we need to tell to eclipse each phase if for example i want to run dead oil case ( with no free and dissolve gas ) i will put only OIL WATER
WELLDIMS - maximum number of well and wells in one group
EQLDIMS- amount of unique equilibrium regions ( in other words regions where you have different contacts and PVT properties)
TABDIMS- this keyword gives amount of saturation regions , amount of different pvt regions, amount of FIPNUM (fluid in place regions) , and maximum number of nodes in pvt and sat tables
number of nodes can be described in following example
For ex i put 2 for nodes in pvt table
but if i supply 3 rows of table , eclipse will read only 2 rows of my table , so put maximum number to nodes ( 100 for example)
AQUDIMS - dimension and type of my aquifer
FAULTDIMS - dimensions of my fault array if present
MESSAGES - This keyword can be used to reset the print and stop limits
for example by default ran stops when errors exceed 10 . You can manually assign number of error messages in 11 parameter of messages keyword
2.GRID
In eclipse you can write data directly in data file or use Keyword include to make link to required information
as grid data is big arrays so we will use include to link
In grid section you need to supply
Grid_array
Actnum_array
Fault_data
Poro_array
Ntg_array
Permx_array
PermY_array
Permz_array
Again it is minimum you need to supply
NOTE!!!!!!
the INCLUDE file should be in the directory of data file. in this case you will have a mess apart system files that eclipse create during run. to avoid this mess instead of standard file link
INCLUDE
'actnum.inc' /

use following construction

INCLUDE
'../GRID/actnum.inc' /

in this case you can create folder GRID and put actnum.inc file directly there
as show in the picture below


GRIDFILE- this keyword tell eclipse to produce EGRID file this file required by schedule and also needed for petrel to visualize results of run
INIT - this keyword tell eclipse to produce INIT file which contains a summary of data entered in the GRID, PROPS and REGIONS sections.

3.EDIT(OPTIONAL)
in this section you can apply modification to grid like multipliers etc lets pass for now this section.


4.PROPS
in this section you give eclipse PVT , SCAL and rock data data
5.REGIONS
in this section you give eclipse all geometrical data described in chapter 1 or 2
6.SOLUTIONS
this section tells eclipse pressure distribution , actual contact position
in our example the will be 4 regions
7.SUMMARY
This section tells eclipse what parameters it should save at each step of run to be then use for analysis like well production rate , etc
8.SCHEDULE
this section tells eclipse the complication of wells and history or forecast rates we will come back for this section later in separate chapter
NOTE!!!!!!!!!!!! for test run when you are on step of preparing and tuning your sim model you can use dates keyword to as eclipse to calculate model for specified period. This approach useful to see if the model stable or not.
EXCERSIZE - Read all keywords description in Eclipse reference manual
Tomorrow i will put test model

[link Point to another website Only the registered members can access]

[meyssam1983]

Thanks a lot

[temr]

So test sim model is ready
in the data folder is Egpet_net.xlsm file with macros
to enable macros working you need to go to security settings in excell and allow macros on you pc. This macro prepare data file out of you excell sheet
Project ready you need to run it is NOSIM mode
Once you run it go to the data folder and open file .prt
prt file will be described on next section
This model is ready all u need to do is comment keyword NOSIM and run macro

[temr]

Sim model ready no schedule

[link Point to another website Only the registered members can access]

[temr]

Replace file in dta folder with this one and run case

[link Point to another website Only the registered members can access]

[temr]

Chapter 5
In this chapter we will review results of run they are
1. PRT file
2. RSM file
3. Restart file
3. Loading results to Petrel
So once we run the case your data folder should consist of following files

There to files which can give as information om runs.
PRT file is a debug print eclipse file where eclipse sore all information on run, messages, errors pass it for now it will be discussed later.
RSM file is the file where eclipse store all information on mnemonics declared in summury.inc
Eclipse mnemonics can be divided on several groope depending on first letter of mnemonic
1. F**** - filed level , like FOPR means field oil production rate
2. G**** - groope level
3. W****- well level
More detail you can in SUMMARY section overview of your eclipse reference manual.
NOTE!!!!!!!!!
form PRT file you can get debug information for run tunings
from RSM file you get all detail information on dynamic parameters like pressure, saturation, rates, injection etc.
To process RSM file use excel as it fast comparing to GRAF
As i mentioned previously for vizualization you have to 2
1 use Floviz - old slow staff
2 use Petrel
For Importing run result data to petrel do following.
Open test project in petrel and go to model section

Highlight Property write click on mouse and choose import on section


In dialog box choose

And find your data file

Click open you will get

Click ok
Now you get new folder

were you have all steps dynamic properties including Pressuer and Satuarion

From this you can easily create any visual data and crossection
Next section will be schedule

pictures

[link Point to another website Only the registered members can access]

[reservoir_engineer]

thanks,
go ahead, we are following.
regards,
reservoir_engineer
_____________________

[reservoirengineer]

temr

Thanks.