Boundary file requirements

A Google Earth file is read literally. The application takes the closed polygon you drew as the plant boundary, treats every polygon inside it as ground to keep clear, buffers every line inside it into a corridor — and decides which kind of feature each one is from the name you gave it. Nothing is inferred from shape, size or imagery.

That makes file preparation the highest-leverage half hour in the whole design. A correct file lays out in one run. A file with a pond named P3 lays out with module tables in the water.

The same rules are available inside the application, from the ⓘ button next to Browse….

Screenshot pendinginputs/kmz-help-dialog.png

Boundary preparation guide

What it shows

The dialog opened by the ⓘ button, scrolled to show the numbered sections and the quick-reference table at the bottom.

How to get there

Click the ⓘ button next to Browse…. Capture twice if needed — the top sections and the quick-reference table.

The plant boundary

Everything else in the file — water, buildings, transmission lines — must lie inside one of those boundaries. A polygon fully contained within a boundary becomes an obstacle in that boundary. A polygon outside every boundary is not subtracted from anything, and the layout will happily cover the ground it sits on.

Screenshot pendinginputs/kmz-google-earth.png

A correctly prepared boundary file

What it shows

Google Earth with the Places panel visible, showing one plant boundary polygon that contains a water polygon, an obstruction polygon and a transmission LineString — with the feature names readable in the panel so the naming rules are obvious.

How to get there

Open a prepared demonstration file in Google Earth Pro — not the application's own file chooser. Expand the Places tree so every feature name shows.

Callouts to add

Label each feature with its type: boundary, water, obstruction, transmission line.

Preparing the file

How features are classified

Classification is by name keyword, and the match is case-insensitive. If a name contains one of the keywords below, the feature is treated as that kind. If it contains none of them, a polygon is treated as a generic obstruction and a line is still buffered into a corridor — nothing in the file is ignored for want of a recognised name.

Water bodies are recognised only from polygons you draw in the file. The application does not look for water in satellite imagery, so a pond you did not draw does not exist as far as the layout is concerned.

Names that work, and names that do not

The keyword can sit anywhere in the name, in any case, with anything around it. Survey codes on their own never match.

Because the match is on a fragment, a name can pick up a keyword you did not intend. Drainage pond carries a canal keyword and a water keyword at the same time; Store near HV yard carries a transmission line keyword. Read your own names against the three lists once before you save the file, and rename anything that touches two lists.

Water bodies

• Draw a closed polygon around the water, inside the plant boundary.
• Name it with a keyword from the water list.
• The polygon is subtracted from the usable area, so no table is placed over it.
• It is drawn in blue on the plot, and exported to the water layer of the CAD drawing and to the exclusion zones of the Google Earth export.

Draw the polygon at the water's maximum extent, not the extent on the day the imagery was taken. The application excludes exactly the shape you give it and adds no margin of its own — if you want a bund or a freeboard allowance around a pond, draw the polygon that much larger.

Obstructions and buildings

Anything that occupies ground but is not water is a generic obstruction: existing buildings, a substation compound, a temple, a graveyard, a rock outcrop, a borrow pit.

• Draw a closed polygon, inside the plant boundary.
• Name it without a water keyword. That is the whole rule — the absence of a water keyword is what makes a polygon a plain obstruction.
• The polygon is subtracted from the usable area and drawn in red on the plot.

If you discover an obstruction after the layout has been generated, you do not have to go back to the file: obstructions can be drawn directly on the plot. See Obstructions and exclusions.

Transmission lines, canals and roads

These are line features, not polygons. A line has no width in the file, so the width of the corridor kept clear around it comes from somewhere else — and which "somewhere else" depends on whether the name marks it as a road.

• Draw the feature as a line through the site, inside the plant boundary. Drawing it as a thin polygon instead means it is read as an obstruction polygon and gets no corridor at all — only the strip you drew.
• Name it with a keyword from the transmission line, canal or road list.
• Every line becomes a corridor. A line whose name matches no list at all is still buffered; it takes the transmission line setback, in the absence of anything more specific. A line you forgot to name does not quietly lose its keep-clear area.

Transmission lines and canals

A transmission line or canal is buffered by the setback on each side, so the total corridor kept clear is twice the setback. The setback is the Transmission line corridor field in Site Parameters:

Raise it for higher-voltage lines. Set it to match the right-of-way the utility imposes on your site, since that is a statutory figure rather than a design choice, and the one setback applies to every non-road line in the file — so a site with both a 220 kV line and a small drain gets the same corridor around both.

Two drawing habits save trouble:

• Draw one continuous line across the site, rather than several short segments end to end. A single line gives one continuous corridor.
• Run the line right out through the boundary edge. A line that stops short of the fence leaves a wedge of usable ground at each end, and tables get placed in it — directly under the conductor.

Roads carry their own width

A line whose name matches a road keyword is not treated as a transmission line, and the corridor setback does not apply to it. A road is buffered by half its own width on each side, so the strip kept clear is the road width — not twice it.

The width comes from the file, not from the input panel. Give the road placemark an extended-data attribute named width, road_width, roadwidth or lane_width, in metres, and that is the width used. A road with no recognised width attribute is laid out at 5.0 m, and any value below 1 m is raised to 1 m.

Two identically drawn lines therefore clear different ground. On the shipped defaults a line named Transmission line 1 clears a 30 m corridor — 15 m each side — where a line named Access road 1 clears 5 m. Name every line deliberately: an access road that picks up a transmission line keyword takes the full corridor out of the layout, and a transmission line named as a road keeps almost none of the right-of-way the utility imposes.

A road is assigned to whichever plant boundary contains the middle of the line, or to the nearest boundary if no boundary contains it. Draw internal roads well inside the plot they belong to on a multi-plot site.

Multiple boundaries in one file

One file can carry as many plant boundaries as the site has plots, and all of them are laid out in a single run.

• Each boundary becomes its own plant. It gets its own row in the summary table, its own area and capacity, and its own folder in the Google Earth export. A TOTAL row aggregates every plant.
• A feature belongs to the boundary that contains it. Water bodies, obstructions and lines are read per boundary, so a pond drawn inside plot 3 is subtracted from plot 3.
• Boundaries must not overlap. Touching along a shared edge is fine.
• Draw every plot, even the small ones. Adding a plot later means editing the file and loading it again, because the boundaries come from the file and cannot be drawn on the plot.

What follows from a multi-plot file — one main control room, a unit substation in each of the other plots, and medium-voltage cables between them — is covered in Multi-plot sites.

What each feature does to the layout

Once the file is read, the features are used in a fixed order:

1. The latitudes and longitudes are converted to metres in the site's local UTM zone.
2. Each boundary is shrunk inward by the Perimeter road width to give that plant's usable area.
3. Water bodies and obstruction polygons are subtracted from it.
4. Every line feature is buffered into a corridor — by the setback on each side, or by half the road width on each side for a road — and that corridor is subtracted too.
5. Excluded terrain is subtracted, when topography is enabled.

Module tables are then placed only where they fit entirely inside what is left, which is why a polygon drawn slightly too small still leaves tables overhanging the real obstacle, and a polygon drawn generously does not. See How placement works.

Quick reference

Common mistakes

Two of these are worth expanding.

Overlapping boundaries are the most expensive mistake in the list, because nothing fails. Each boundary is laid out as its own plant, the overlapped strip gets tables from both, and the total row of the summary table reports a capacity the site cannot physically hold. It is easiest to catch by area: check each plant's acreage in the summary against the survey before you trust the total.

A boundary drawn as a line is the most common export mistake, because some drawing tools default to a path when you click around a shape and only close it if you use the polygon tool explicitly. Check in Google Earth that the shape has a fill, not only an outline, before saving.

When a boundary fails validation

If a ring is not closed, or is otherwise not a usable polygon, loading the file opens the Boundary Validation Issues window. It lists every problem boundary with a description of what is wrong and a checkbox against each.

Screenshot pendinginputs/boundary-validation.png

Boundary validation issues

What it shows

The dialog listing each problem boundary with its description and exclusion checkbox, plus the buttons at the bottom.

How to get there

Load a file containing at least one boundary whose ring is not closed.

Callouts to add

Outline one row's checkbox.

You have two choices:

• Exclude the problem boundaries and proceed. Tick the boundaries you want to drop and continue. The remaining boundaries are laid out as usual. Useful when one plot of a twelve-plot file is malformed and you want to see the other eleven now.
• Cancel and fix the file. Nothing is loaded. Correct the geometry in Google Earth, save, and load again.

Where to go next