Distributed Data Processing Platform

/Distributed Data Processing Platform
Distributed Data Processing Platform 2017-01-13T17:16:55+00:00

Silixa has developed the industry leading distributed temperature and acoustic sensors capable of enabling a range of in-well applications such as seismic evaluation, production, injection and stimulation monitoring and well integrity surveillance.

Silixa’s DTS and iDAS™ systems can operate simultaneously on separate fibres in the same well giving parallel data sets which, when suitably integrated, processed and visualised, can create high confidence well based answers.

To facilitate true integration of distributed temperature and acoustic data Silixa have built a platform software called  Distributed Data Processing Platform (DDP Platform). The DDP Platform has an architecture that allows new processing functions to be added to a baseline in the form of plug-ins. The baseline functionality provides all of the tools that the user needs to access and load Silixa temperature and acoustic data to perform depth and time based correlations and to plot the data side by side.

The baseline also allows for the conversion of acoustic data to the frequency domain by FFT and for differential temperature computation.

Silixa has a number of proprietary processing methods for acoustic data and continue to develop new methods of extracting high value information from large acoustic data sets, as and when these processing functions are defined they can be written to a plug-in format and imported directly to the DDP Platform to expand the functionality of the software without the need to replace the baseline. An example is the computation of speed of sound and flow speed from propagating acoustic signal.

All processed data is by default saved to a proprietary format readable by the DDP Platform; however, an additional plug-in allows for data to be exported to other formats such as HDF5.

A data viewing environment within the software allows the user to manipulate how the temperature and/or acoustic data should be presented and then to save data plots to file for each user-defined time window. Plotted data can also be exported to Log Ascii Standard (LAS2.0).

Processing Routines

On opening the software the user is presented with a list of processing functions on the left-hand pane as shown in figure 1. As new functions are added in the form of plug-ins they can be imported from the ‘Routine’ selection on the task bar and will then appear as an available function.

To build a complete processing routine the user will drag each sequential step to the right hand pane. Each function contains specific settings to define its operation which the user can adjust. A partial or completed routine can be saved to file so that it can be retrieved from a library.

The routine will typically begin with the setup of data file paths; a routine can operate on both DAS and DTS data simultaneously. As the routine is running a console, see figure 1, continuously provides text updates on the process to allow the user to follow progress.

The output of processed data is then saved continuously to a proprietary format which can be used to create used-defined plots calling on the results of both temperature and acoustic data. A processing routine can be broken down into branches, each one saving to a separate output data file. This ensures that data can then be accessed and plotted with a greater degree of flexibility as it is not necessary to reload all processed data to view sections of interest.

Depth and Time Correlation

All temperature and acoustic data is depth and time stamped at acquisition.

When field data is loaded it may be necessary to perform depth corrections and/or time shifts; an example would be moving a data set from local time to UTC or vice versa.

The DDP baseline provides functions to apply time shifts to entire data sets or to individual files, similarly acoustic or temperature data can be depth corrected by applying both offset and stretch corrections.

The reconciliation of time stamps allows for acoustic and temperature data to be plotted side-by-side ensuring that the data originates from the same window in time and space.

Figure 1: DDP Platform User Interface

Basic Processing Routines

The DDP  Platform baseline allows for basic temperature corrections and re-averaging as well as the computation of temperature changes over user-defined periods or from the geothermal, or other, baseline. Time lapsed temperature data can then be plotted as sequential curves or as a colour VDL (waterfall).

Acoustic data can be normalised/averaged by channel or depth and can be converted to the frequency domain by FFT over user-defined time windows. This allows for acoustic data to be presented as a conventional waterfall plot or as a frequency distribution VDL. Acoustic RMS values can be computed for user-defined frequency bands and presented as individual curves. Figure 2 shows an example of this type of temperature and acoustic data processing and visualisation.

The baseline also allows for consecutive FFT results to be stacked in order to improve the SNR of a given acoustic feature which can be used to extract information from quiet environments.

Advanced Processing Routines

The plug-in architecture allows for new processing functions to be imported to the DDP Platform to allow for more complex routines and answer products.
Examples of available plug-ins:

  • Zonal production allocation from RMS amplitudes
  • Zonal production allocation from rule based temperature analysis
  • K-omega: speed of sound and flow speed computation
  • Event detection
  • User defined computations

The architecture allows for new routines to be quickly developed in the appropriate language and subsequently loaded into the DDP Platform. A new user of the software could request a particular routine to be added allowing for a new processing step to be applied to the acoustic and/or temperature data.

Figure 2: Basic data presentation

Track 1 Inclination angle
Track 2 DTS temperature data & TVD
Track 3 Differential temperature (10 minute, 30 minutes, and 60 minutes)
Track 4 RMS frequency bands
Track 5 Frequency distribution
Time Track Imported time-based well pressure and external probe temperature data (BOTTOM)

Data Visualisation

On accessing the software the user can select whether to access the processing environment or the viewing environment, where processed data files are already available, or directly after a processing sequence, the user can enter the viewer setup screen and define the layout of the data for visualisation. Figure 3 shows the viewer setup screen.

On the left hand pane the user has access to all available data files which have already been processed and path-selected. Each file will represent a different type of data and so can be dragged into the appropriate graphing tab on the right-hand pane. The user then has access to the highlighted data file to manipulate how the data should be presented such as curve colour and style, axis range and labelling.

Static data such as imported depth based petrophysical logs (eg. GR) and time based well data (eg. pressure) can be imported via the main menu and placed accordingly in the graphing tabs.

The well completion can be constructed by the user from a function accessible in the main menu and can be saved to file. This will place marker lines and labels overlaid on the graphs to highlight the position of key well and completion features.

Once the user is satisfied that the presentation set up is complete they can move to “view data” and be presented with the view as shown in Figure 2. The presentation template can be saved and reloaded so that similar results can be repeated.

Since the majority of distributed data will be acquired over a period of time, and the processing steps will usually focus on smaller time steps within that period the visualisation result will typically include multiple plots each representing a subsequent period in time (eg. 1 minute), it is critical that the temperature and acoustic data and the processed answers are accurately aligned in time.

The horizontal track on the output plot is time-bound and can show time-based client data such as bottom hole gauge pressure or temperature. The vertical blue line (See Figure 2) displays the point in time from which the presented data is taken. This allows the user to correlate the distributed data set and answer products to other well performance indicators, as well as correlating events to depth-based petrophysical and well construction features from track 1.

The viewer screen (Figure 2) allows the user to step through the data in time and to view the changing conditions, in doing so the user can elect to export the plots to an image file (.PNG) so that they can be used in other applications (Word, Powerpoint) or can be animated into a short video which adds the dimension of time to the data set visualisation. The user can also elect to export the content of the plot to LAS2.0 format within this viewer environment.

Figure 3: Viewer Setup interface