Skip to content

Any Asked Questions

Jump to bottom
Mathew Topper edited this page Oct 24, 2018 · 2 revisions

This page contains a list of answers to questions from the user community.

The DTOcean version 1.0.0 installer fails with the message "DTOcean installation directory not found!". What do I do?

This is a sporadic bug with the installer. Please see DTOcean Installation Solution for a workaround until the next version is released.

I've installed the DTOcean application but I can't access the database. How do I install it?

The database is installed separately to the main DTOcean application and requires a PostgreSQL installation. Full instructions can be found in the download. Once the archive is unzipped follow the instructions in the README_database_installation.txt file.

I'm following the video tutorial #1 part 3: "Input Preparation", how do I select the "shetland" database example?

The database in the video tutorial is not the same as the database released with the tool. For instructions regarding the examples in the final database, please see the README_example_data.txt file in the database download.

The video tutorials will be updated in future to include the examples in the distributed database.

Do the lease area, cable corridor and cable landing point have to be contained entirely within a site area? Or can the Site area share points/boundaries with a lease area or a cable corridor?

The site area is a vanity variable that is used to show the user whereabouts the site is when selecting between sites. It is not used at all in the design calculations. It is only useful for the site area to include the lease area and cable corridor boundaries for some of the plots.

The lease area and the cable corridor requires overlapping points between these two bathymetries although the resolution of the cable corridor does not necessarily have to match that of the lease area bathymetry. The boundary definitions just have to demarcate sufficient points in both bathymetries to reveal overlapping points.

Can a cable landing point be anywhere within the cable corridor, or must it be close to/on a boundary?

The landing point does not have to be defined directly on a bathymetry point, but the electrical module, for instance, will snap it to the nearest point in the filtered bathymetry, so it's probably worth defining it at a known point within your chosen domain.

In the beta.bathymetry table, are we defining the points for the lease area, or the site area? If for the site area, does this table contain points that are inside the cable corridor, or not?

The bathymetries are defined in four tables. For the lease area these are beta.bathymetry and beta.bathymetry_layer. For the cable corridor they are beta.cable_corridor_bathymetry and beta.cable_corridor_bathymetry_layer.

An explanation of what should be contained in these tables is in section 9.4 of the technical manual. Note that there are a lot of repeated and unused columns within the database tables. Presently there isn't time to describe everything that is and isn't used, but we will endeavour to produce a "map" of the used tables and columns in the future.

Do bathymetry points have to be within the boundary, or can they be on the boundary line? Or can we define bathymetry points outside of the boundary without issues?

The purpose of the lease and cable corridor boundary polygons is to tell the database which points from the bathymetry tables should be "passed" to the main DTOcean application. The selected points are actually copied to the related "project" tables from which they are then read by the tool. All of the site and device server side filtering works in this manner.

The server side filtering will also collect points which lie coincident to the boundary. It is possible to define bathymetry points that lie outside of the stored lease and cable corridor boundary definitions, but it is perhaps inefficient to do so.

Does the resolution for the bathymetry data points matter (i.e. 12m vs 100m)? Does the resolution have to be constant (i.e. a point every 12m)? Do the points have to be in a Grid pattern?

Grid resolution in the lease area is a balance between accuracy and performance. Some common sense must be applied as the design modules do not interpolate between points, thus if you want to lay a cable between or build foundations for devices then a 100m resolution is unlikely to be sufficient. There is no correct value for this, simply a choice of the modeller.

The situation is less important in the cable corridor where only a cable need be laid. Again, higher resolution will provide a better answer at the cost of computational time. The same holds for the extent of the cable corridor which will increase computational time as the area increases.

The resolution of both grids must be constant and in grid format but the resolution of the lease area and the cable corridor need not match. They must have overlapping points, as explained in the previous question. Further information is given in section 9.4 of the Technical manual.

In the example database, some geometry data has been inserted as latitude/longitudes (such as site_boundary in beta.site), and other geometry data has been inserted as UTM (such as lease_boundary in beta.site). Is there a reason for this?

In general the tools work with a grid based coordinate system (in metres) which is local to the array lease area. UTM is the most well known of these but any similar projection could be used as long as it is described in the beta.site.lease_area_proj4_string column as a valid Proj4 string.

Some other values are in lon/lat coordinates to allow referencing outside of the local coordinate system. One of these is the site boundary polygon which is used to select between different sites and therefore lon/lat coordinates are the most appropriate. If not in lon/lat all coordinates should be given in the local UTM-like system.

Please refer to section 9.4 of the Technical manual for some further details.

Clone this wiki locally