KAPPA Server Guided Session #1 · Before proceeding in the tutorial itself, ... temporarily launched from Ecrin, ... Doc v5.0.02 - © KAPPA 1988-2013 KAPPA Server Guided Session #1

Kappa Server v5.0.02 – Doc v5.0.02 - © KAPPA 1988-2013 KAPPA Server Guided Session #1 • KSGS01 - 1/38

KAPPA Server Guided Session #1

A. KAPPA Server overview

Before proceeding in the tutorial itself, it is necessary to go through the description of the

KAPPA Server architecture and organization.

KAPPA Server is a server solution installed on a dedicated machine, that permanently mirrors

the raw data (Historian database) for fast processing, reduces the number of points with

wavelet based filtration, stores and shares the filtered data and exports filtered data sets to

third party databases.

The KSC (KAPPA Server Controller) is used to configure the settings of Diamant Master

(mirroring, filtering, users’ privilege).

The KAPPA Client module, temporarily launched from Ecrin, operates KAPPA Server and

transfers the data and the technical objects between KAPPA Server and the Ecrin analysis

modules.

The KSC can also be used by the client to monitor on-going processes in KAPPA Server

(mirroring process, filtering process).

B. Introduction

The objective of this session is to illustrate the KAPPA Server key features, seen from the client

side. You will be guided through the startup of a project and the exploitation of the Permanent

Gauge Data.

This will give a simple workflow using the Ecrin / KC / KS integrated environment to transfer

data, technical objects between the mass data storages and the analysis modules.

To practice the use of Data Transfer from Raw Data sources definition it is necessary to have a

'Database'. In order to avoid any 'wrong operation' that could damage the existing real

database we will indicate the procedure to 'simulate' a raw data source from the server with

BLI files (BLI stands for Binary Large Indexed).


C. Requirements

The necessary BLI files must be downloaded from the KAPPA website:

http://www.kappaeng.com/downloads/default.aspx?s=6&t=s

That can only be done by a registered User, with the proper Username and Password.

Select 'Additional Tutorial Data' to download the zip file containing Well #1-EX-

PDG01p_002.kbl; Well #1-EX-PDG01q_002.kbl.

These BLI files contain pressure and production data, one value every 5 seconds during 7

month. That represents more than 3 million data points for each channel.

D. Configuration

D.1. Server configuration

This part of the configuration has to be performed by the IT engineer in charge of the KAPPA

Server.

Open an internet browser and type the server name in the address bar.

Run the KSC (http://Server Name/ksc) on the server, using the Admin account (created

during the installation) to add a user (the client).

Note: We strongly recommend that you assign rights with the 'Create' privilege only, at least

while the new user becomes familiar with KAPPA Server. This is to avoid potential mistakes on

the existing data already created by actual users. The privilege can later be extended.

http://www.kappaeng.com/downloads/default.aspx?s=6&t=s


Copy the BLI files in a folder visible from the server or on the server.

i.e.: for the guided session, let’s call it: 'C:\Training Database\bli'.

Define the data source through the KSC. The Administrator privileges are required.

In the 'Data sources' page of the 'Connections' section, select the 'Bli Plugin' option.

After pressing 'Create', define the data source, that you may call 'Training BLI' and indicate

the relevant BLI directory path as illustrated below. Hit 'save' to confirm. The data source

name must be written in exactly the same way as was defined when copying the folder to

the server.


Your data definition appears then in your available definition list. The status will only appear

green once the server has tested the data.

For more information about the Plugin type selection and the Data Source Definitions see also

the KSC on-line help, by clicking on the blue top-right question mark.

D.2. Client configuration

Once the IT engineer has installed and configured the KSC on the KAPPA Server, the KAPPA

Client users must install and configure Ecrin on their workstations.

Once Ecrin has been installed, start Ecrin and from the 'File' drop-down menu, select

'KAPPA Server connection...'.

Add a server connection.


The available servers are displayed with their corresponding privilege levels.

Your server should appear with the icon , indicating that it is started and available.

Select your server and hit 'Connect'.

Kappa client will automatically be installed.

Once connected, the field list is displayed and shows if any, the already existing fields.


E. Creating a field and loading data

E.1. Creating a field

When launching KC for the first time, no field should exist.

To create a field, press on to proceed and input the field name as indicated

below. Hit 'OK' to confirm.

The field called 'Training field' is created on the KS. It can then be opened and accessed by any

other declared user.

Open the 'Training field'.

E.2. Loading the pressure data

Create a well by pressing . You can input well description in the 'Properties' dialog. Let

all by default. Press ok. The well icon is on the tree.

Create a Production by pressing and validate OK for the default name ‘Production #1”.

The production #1 node is in the tree creating oil, gas and water.


A right-click on the well #1 icon displays a

pop-up menu which offers multiple facilities.

Depending on the size of the data set, we can load data either by choosing:

Add new gauge ...

Or

Add new data ...

For a large size data set we must load the data as 'New Gauge'. It presents the advantage to

create a mirror of the data set and to display only a reduced number of points (around 1/1000

points).

It makes the large data set handling easier and faster: as often as needed, Filtered Data can

be created under a ‘Gauge’ node with the required number of points according to the analysis

needs by adjusting the filter settings.

With a reduced number of data, we can load the pressure as 'New Data'. The complete set of

points whatever its number will be displayed entirely. In case of large amount of data set the

'Add new data' option should not be used as the display would be extremely slow.

Note: right-clicking on one of the Production icons displays a pop-up menu only allowing to

load complete 'Data (no filter)' type files. The procedure to load filtered production data will be

indicated below.

Select 'Add a new gauge'.

In the ‘Define data source’ dialog, activate ‘Database’ and select the 'Training Bli' created

by the IT engineer.


Press 'Next'.

You are then connected to the 'simulated database'.

In the 'Load tag(s)' dialog press on to display the available tags.

Well#1-EX-PDG01p_002 contains the pressure data.

Select Well#1-EX-PDG01p_002, then press on and to display

the main statistics of the pressure channel.

Press to proceed.


The adequate reading format is already defined since it is included in the database.

Press to initialize the mirroring process.

A message pops up to indicate that you will need to select the ‘update option’ to see the

progress locally. Click OK.

In the browser tree, select the gauge .

Click on to view the gauge info if it’s not selected by default.

It may take a few minutes before you can see data in your gauge info summary. If you do not

get any data, click on the 'Update' button. If after several attempts, nothing happens,

contact your Diamant Master Administrator.


The total number of points is 3 377 929 but only 3379 points have been selected to be

displayed in the preview. The preview settings are under Ecrin /settings/Loading data and the

preview data are 1 data every 1000 points and cannot be changed. The preview mode is

automatically applied for all loaded data.

Click on to display the plot. Only the 3379 preview points are plotted.

Clicking on the 'Update' button updates all the plotted Gauges and Filtered data with the

KS current content. However, it does not imply the synchronization of the KS with the raw data

sources.

Pressing on the button synchronizes the display in KC with the KS content, which means

that it adds and updates any item created or modified externally (i.e. by another user) and

that could not be updated automatically on the client by the KS.

The mirroring update process (from raw data to the BLI) is controlled by the 'Processes' page

of the KSC. By default, it is set to be triggered every two hours.

E.3. Getting details on the mirroring process using the KSC

The KSC (KAPPA Server Controller) can be launched from the client station for monitoring the

various processes.

Input http://your_server_name/KSC/ in your Internet Explorer.

The 'Summary' page displays the information about the server, the users, the field and data on

the KSC and also indicates the status of the various processes.

http://your_server_name/KSC/


The ‘Tasks’ page can be accessed only with the Administrator privileges.

It allows to visualize and to modify the KSMP (KAPPA Server Mirroring Process) settings.

The Timer value is the interval between two Updates of the mirror with the 'new' data from

the 'Raw Database'.

The Max Record (i.e. 10000 pts) is the number of downloaded points during each query.

If various gauges are to be updated simultaneously that will be done, gauge by gauge,

by subsets containing 'Max Record' points (i.e. 10000 pts).

The 'Mirroring tasks' page shows the progress of the mirroring of all the gauges, the current

number of mirrored points, the number of new points to be mirrored. It gives the possibility by

right-clicking on the icon to Stop or Start the mirroring process of any gauge.


E.4. Creating a filtered data

Even if the gauge (mirror) contains the full set of data points, this large number of data is not

adequate for analysis. It is then absolutely necessary to filter these data to get a reduced and

representative subset of points for analysis purposes.

The filter reduces the number of points, respect the information and have a frequency adapted

to the needs (i.e. PTA or PA). All these properties are obtained in the same process, based on

a wavelet algorithm, acting as a high pass filter around the pressure breaks and acting as a

low pass filter for the rest of the data.

The filtering process includes a first denoising phase and then a post filtration where the

number of points is reduced according to criteria, versus time and pressure change, adapted to

the objective.

This filter typically divides the number of points by 100 without losing significant information.

Right-click on the gauge ,

a pop-up menu is displayed.


Selecting 'Add new filtered data' leads you to the filtering setting dialog where you can reduce

and filter the raw data (see also 'Filtration' in the KC on-line help).

Accept the default settings, except in the filter decimation settings: set the Delta T to 2hr

and the Delta Y to 10 psi, then, Apply and OK.

A new ‘Filtered data #1’ is created in the tree under the gauge node.


Click on to display the filtered data statistics. The number of pressure points has been

reduced to approximately 9308 points (depending on your filtering settings).

If you go back to KSC in the Internet browser, selecting the 'Tasks/Filtering tasks' window you

can see the current number of filtered data points.

Press to display the plot page.


By right-clicking on the node of the filtered

data ,

a pop-up menu is displayed as illustrated next >.

The filtering process automatically updates the data every 100,000 new points present in the

gauge.

If you need to work on and process a set of filtered data, you need to 'lock it'.

So click on Tab ‘Edit’ and lock it by clicking on ‘Start editing’ . It will lock the data set in

order to prevent other users to modify it simultaneously.

When you finish modifying the data channel you can unlock it by pressing the icon .


E.5. Loading a filtered production data

As previously stated, we cannot directly load unfiltered data in the production channels.

The correct procedure is:

1. Map the complete production data as a 'new gauge' under the 'Well#1' icon.

2. Add filtered data below this 'production gauge'.

3. Drag and drop the filtered production data into the production icon.

Select 'Add a new gauge' from the Well #1 pop-up menu.

In the Data source dialog select the 'Training Bli' created by the IT engineer. Hit 'Next'.

You are connected then to the 'simulated database'.

In the 'Tag Search' dialog press on to display the available tags.

Select , then press on and to display

the main statistics of the production channel.


Press to proceed.

Change the data type from ‘Liquid rate’ to ‘Oil rate’.

Press to initialize the mirroring process.

In the browser tree, select the gauge .

Click on to view the gauge info.


The total number of points is 3 164 501 but only 3166 points have been selected to be

displayed in the preview.

Click on to display the plot. Only the 3166 preview points are plotted.

Once the data are loaded, right-click on the gauge and select 'Add

new filtered data' in the pop-up menu.

Accept all the default settings and proceed with the filtration process.


Click on to display the filtered data statistics. The number of production points has

been reduced to about 32453 points.

Click on then drag & drop the filtered data onto the Oil

production icon .

If the drag and drop is not allowed, that means that the data type was not changed

previously to Oil Rate, you can change it now by double clicking on the rate filtered data

node.


Change the aspect of this data set to a green line by right clicking on the filtered data and

changing the ‘color’ to green and ‘type’ to solid in the ‘aspects’ tab in properties.

Click on to display the plot.


E.6. Shut-in identification

A number of Build-ups are visible on the pressure history. Yet the rate history of Well#1,

loaded from the raw downhole permanent sensor is unlikely to present clear null sections

synchronized with the pressure history. In order to ensure synchronization between rate and

pressure for analysis, we will just tag the existing build-ups as such.

E.6.1 • Shut-in indicator

Select the EcrGS01p filtered pressure data node in the data tree (in Plot view and not in Table

view); select the ‘Shut in ribbon’:

Press on Start editing in order to lock the channel and to be allowed to work on it.

The first icon, can be used to set a time selection as a build-up. This is the manual

selection tool.

The second button runs a semi-automatic option to define build-ups. The principle of

this option is to pick any points within a build-up, and to let KClient find the start and end

times automatically. After enabling this option just click anywhere in the period and the shut in

will be defined.

The third button is a fully-automatic option to define shut-ins periods. This time the

algorithm will automatically find all the shut-in periods for you, based on a minimum duration

and delta p settings.

For the moment, let us select the longest shut-in as shown, using the semi-automatic function

click anywhere in the build up to define it, then click on and pick it again to select

it for use:


Detected build up

A node ‘Shut-in indicator’ will appear in the browser. This is a logical function that is 1

when a shut-in is detected, and 0 anywhere else.

E.6.2 • Shut-in express

Once a build-up is tagged within the build-up indicator channel, it is possible to get a quick

look at the corresponding loglog plot in Saphir with just a few clicks.

A shading appears on the build up, since we have selected it.

Click the ‘Shut-in express’ icon ; this is the last option of the ‘Shut-in’ toolbar.

A dialog is displayed to control the transfer of data to Saphir.

Shut-in express called for the main build-up


Through this option, a Saphir file will be built on the fly with the current pressure data, and

‘some rate history’ which will be 0 over the shaded zone (the selected shut-in period). For the

rate history prior to the build-up we need to select a given fluid (here only Oil is available) and

we need to select among:

The actual production = ‘OIL Rate’ channel under Well#1.

Part of the actual production only.

A manual entry of production time tp, and rate Q.

We will choose the first option – select ‘Use the total production history’ in the dialog. Leave

the ‘truncate data at the end of the last shut-in’ checked so that only data up to the selected

shut-in period is automatically carried over to Saphir.

Note: the rate history currently holds more than 26000 rates. When we go into a real analysis

later we will reduce those rates to something smaller in Saphir; for the time being all we want

is to have a quick look at the loglog trend without calculating any model but only time

superposition so this is OK.

Without further inputs, we are taken straight into Saphir after the Extract dP step clicking on

highlighted Ecrin icon.

Shut-in express


If we wanted to carry on with a real analysis, the first step would be to visit the Main

Information section of the Saphir document to enter the proper Test and PVT parameters. We

will not do it in this session.

Close the Saphir file (do not save).

Note: you can select several shut-ins periods and still use ‘Shut-in express’. By doing so, when

in Saphir you will have multiple build-ups or fall-offs extracted at the same time in the loglog

plot and the semilog plot, very useful for comparison.

E.6.3 • Automatic Shut-in

Go back to KAPPA Client and stay in the EcrGS01p Fltd data node. This time we are going to

use the automatic shut-in functionality to define all the build-ups in one go. This option allows

the user to define multiple shut-ins in the history using an algorithm which considers the

pressure behaviour and shape.

Click right on the node, then above the plot window, click on to reset the

channel to 0. You can do this by right clicking on the shut-in Indicator node as well. All the

defined shut-in periods will be erased.

Click on the EcrGS01p Fltd data node and display the plot.

In the tab ‘Shut in’, after clicking on ‘Start editing’ click on the ‘auto’ button .

The shut-in detection settings dialog appears:

Make sure that ‘analyze the signature of the

data’ is checked with the displayed option

selected.

Set the minimum pressure change for a shut-

in to be 200 psia and minimum shut-in

duration to be 1 hour.

Press OK.

All the BUs detected are highlighted according to the criteria. See figure below.


Detected build ups

Clicking on the node ‘Shut-in indicator’ displays the all the BU’s detected.

Before using these multiple build ups for analysis, we need to clean and adjust the production

history in order to synchronize it with the pressure history.

Note: Although the automatic detection is very useful and can significantly speed up the data

pre-processing stage, it is based on an algorithm which may sometimes come up with incorrect

limits if the pressure build-up exhibits such behaviors as irregular humping, soft shut-ins, very

noisy shut-ins, etc. In these situations, you can correct the limits manually or semi

automatically. In our current case, the selection made automatically can be kept.


E.6.4 • Corrected production

Right-Click on the well# 1 node and select

the option .

Keep the default setting in the dialog:

Press OK.

The created ‘Corrected data’ is a pre-defined derived channel (explained below), its equation

can be seen by double clicking on the ‘oil rate’ channel under the ‘corrected production’ node

(expand the node to see the ‘oil rate’ channel):

Derived channel definition


In the ‘Aspect tab’ set the ‘corrected production’ oil rate curve in red.

The objective of a derived channel is to create an output data set resulting from n input

channels. Writing ‘Y’ the Y values in the inputs, the values of the new channel are defined as:

Ynew[t]= f(Y1[t], Y2[t], Y3[t], …, Yn[t],a1,a2,…, ak).

Where the parameters ‘a1’… ‘ak’ are user defined constraints.

In this particular case, the channel is defined in such a way that the resulting ‘corrected

production oil rate’ channel is set strictly to Zero whenever a shut-in is detected in the build-up

indicator channel, and will leave the rate as it is originally in ‘raw production’ anywhere else.

As a derived channel, the corrected production would be extended automatically as the input

oil rate gets updated via the database. All we would need to get this channel up-to-date is to

pick the future build-ups so that the build-ups definition is up-to-date.

Real rates; BU Indicator; Corrected rates

Select the Plot view for Well#1 , and in the ‘Plot’ ribbon click on Show . Select

‘production #1 ‘corrected production #1’ and ‘shut-in’, to view the difference between the

original production and the modified one.


A blown-up view at the beginning of one of the longest build-up is presented on the above

figure. Note that the corrected production is a representation of the surface rates. When

downhole rates are present and we have soft shut-ins of the kind above, it may be preferable

to define the build-ups only at the time when the downhole rate vanishes. This is beyond the

scope of this session.

By applying this formula we can see here that the rate now is strictly zero when the build up

indicator is 1 (means that a shut-in is selected), eliminating thus potential noises when there is

no flow, and strictly cut to zero at the start of shut-in for synchronization with pressure

history.

E.7. Creating Other Derived Channels

Derived channels are the results of any calculations that can be required by the user.

Any data channel can be used in the formula written by the user in the dialog accessed by

pressing in the top toolbar.

The user can select within basic operation types and within standard mathematical functions

by pressing .


Write a simple formula to detect production values smaller than 1000 bpd and set them to 0

and keep the rest unchanged.

Type ‘if(a<b, 0, a)’ in the formula box.

Press on to display the list of the formula parameters:

‘a’ is the production channel values, specify it by pressing on and checking the Well#1-

EX-PDG01q_002#1 Filtered data node:


Check ‘b’ as a constant and input 1000 as follows :

Press OK.

Click on the ‘Derived data’ node and press on to display the plot.

All the production values less than 1000 bpd have set to zero to declare them as build ups.


E.8. Setting alarms

The same process can be use to set alarms to warn the user that a certain function or data

value has reached a certain predefined value or when boolean logic function is set to false.

Let’s set an alarm if the pressure drops below 8600 psia.

Press on to get access to the ‘derived channel definition’ dialog and type in the

formula box: if(a<8600,1,0).

Press on to display the list of the formula parameters:

‘a’ is the pressure channel values, specify it by pressing on and checking the Well#1-EX-

PDG01p_002#1/Filtered data node:

Select the Output type ‘Logical’:

Press on the tab ‘Alarm’ and set the dialog as follows:


Set particularly the type of alarm to ‘Historical’ that allows scanning the past data from the

specified date (23/04/2001).

‘Instantaneous’ type scans the newly imported data.

We can set the alarm in order to send a warning message to the person in charge.

Press to proceed.

At the beginning of the scan the ‘Alarm node’ appears as before any limit value was

detected, then is appears as when the logical function is ‘true’.

The alarm turns to when the logical function was true in the past but false currently.

The logical function can be built for multiple purposes, i.e. for detecting an abnormal noise the

user can consider the difference amplitude between two strongly and a slightly smoothed

filtered data.


E.9. Transferring to Saphir for PTA analysis

Initialize a Saphir session; accept all the defaults values and options.

Open the browser in Ecrin.

You can drag & drop the filtered data from KC to the Saphir 'Tested well' pressure gauge

folder.

Or, after selecting the filtered data in KC, use Kappa send

in the KC ‘Home ribbon’ toolbar to send it to Saphir.

Do the same for the corrected production and drag & drop it into the Saphir production

folder (check before in KC that the Production data type is ‘Oil rate’).

Or, after selecting the corrected production node for the OIL-filtered data channel in KC,

use ‘K Send’ in the KS toolbar by right clicking on the node to send it to Saphir.

Afterwards, close the Browser.


E.9.1 • Start of PTA

Go back to the Saphir main window.

You will now see both the Pressure and Production history plots (revert the Pressure gauge to

the full history if you loaded a partial section as indicated above).

Go to the 'Edit Rates' tab to visualize the detected build ups.


Go back to the analysis tab to proceed with a pressure transient analysis, extract the BU

indicated below, using the 'Extract dP' button .

Then you obtain the following result:


E.9.2 • Partial reload

Build-up data may be insufficient for analysis and we may need to get a more detailed

pressure description.

In KC create a ‘reduced’ new filtered data set using a very strong decimation:

Delta T 12hr and Delta Y 100psi.

Send this reduced set of point to PTA.

After extraction in PTA we get :

The BU data are not sufficent for an effective analysis.


A part of the data can be reloaded with different filtration settings in order to get more points.

In KC, ‘Edit’ ribbon, lock the data to edit it.

Select by time values a subset around the considered BU:

Click on , you can back up or not the original data.

You are back to the 'Filters' tab, select:

The wavelet sampling for Pressure Transient Analysis.

Decimation: dt max 0.2hr, dy max 1psia.

Apply and OK.

The BU period is re-populated:


From the home tab, press on to process the re-populated data.

You get then in Saphir a more accurate loglog plot for the BU analysis.

Once the data are loaded in Saphir, it is recommended to save the file on your C: and then to

include it in the KC field.

From KC, add a folder under your Well.

Select 'Add new file' and then indicate the path to your Saphir file.

Your file is now associated and saved in the KC project .

KAPPA Server Guided Session #1 · Before proceeding in the tutorial itself, ... temporarily launched from Ecrin, ... Doc v5.0.02 - © KAPPA 1988-2013 KAPPA Server Guided Session #1

Documents