Recent changes

Calculation of the tank factor effect for individual systems incorrectly apportioned the kWp part of present cost, causing the optimiser to overestimate the cost ofusing these kinds of individual system. This bug has been removed.

The model's consideration of pumping costs was incorrect; the cost of pumping energy was incorrectly not converted into a lifetime present cost when combined with the cost of heat from supply points. This has had the effect of reducing the cost to the optimiser of supplying heat, especially with a large pumping overhead configured.

The model's estimate of building height is displayed in the editor.

Click the pencil icon next to a network problem in the project page to rename it.

The candidate editor would sometimes erroneously amend buildings' peak demand values, and some improvements have been made to technical details of the optimisation.

Hiding unconstrained candidates no longer makes the application forget any edits the user has made to those candidates.

The demand editor (e key) now lets you adjust demands by a percentage.

If a pipe is marked as existing with a max. diameter, that is the diameter of pipe that will be used. Also, the UI now displays the pipe maximum capacity separate from required capacity for networked demands. This new information will only appear for new model runs.

• Cooling demand model works again - it was accidentally disconnected.
• Entering numbers doesn't make the focus jump around.

A bit more information is shown about runs on the project page.

The spreadsheet download would sometimes produce an invalid spreadsheet when a solution involved counterfactual heating systems.

The pipe costs table erroneously displayed zero for default capacity values, and the spreadsheet import/export now preserves the default civil cost.

Carbon factors for supply plant were not properly incorporated when evaluating carbon limits. Now they are.

When using market tariff, the market rate used is now shown for building that don't get connected, as well as those who do.

Buildings which could not be supplied because they are connected only via another building are now marked as unreachable when solving a problem. Previously, they would render the problem unfeasible for the optimiser.

The documentation has been updated with some information about the demand estimation model.

The spreadsheet download now contains the solution summary page

The help documentation has been updated to reflect new features in the application.

The spreadsheet upload now loads supply model parameters.

The solution summary display now includes an equivalized cost column.

• The selection info panel would incorrectly display a building's peak demand for its insulated demand, if it had not been included in a solution.
• Manually edited buildings or paths whose constraint status is forbidden were incorrectly reset to their original state when they were not visible on the map for a period of time. User edits are now remembered even when you can't see the edited candidate.
• Connection costs have been moved into the pipe costs page, for consistency.
• The checkboxes controlling insulation & individual systems are disabled when not in whole-system mode.

• The supply model incorrectly took the grid offer price as a cost rather than revenue, making a CHP worse value than an otherwise equivalent boiler, rather than better value as it should be.

• The supply model's interface to the network model only considered the first supply location in networks with multiple supplies, so the supply model would be solving for only part of the system peak / annual demand.

  Now, if there are multiple supplies their peak and annual demands are summed for the supply model, which still solves them as a single system. This is still a simplification, but hopefully a better one.

• Projects can now have their own associated LIDAR, used for maps in the same project.
• When creating a new map the heating degree days field will be pre-populated with a relevant value.
• Individual systems can now represent a 'hot water tank' for sizing. On the individual systems page, there is a new field for 'Tank factor'. If supplied, this field is used to compute a new peak for the building when using that individual system as a multiple of the annual demand.

• NPV or annualization terms of zero are taken as one year. Previously this would stop the model running.
• The hot water demand part of the demand model produced values 1.15 times larger than it should have; this is fixed, for new maps only.
• The application should do a better job of processing email addresses when inviting people to a project.
• The diversity calculation is corrected in the case where the optimisation solution includes multiple paths from the supply to a vertex. Previously, if one path had a very small heat flow the diversity on that path could be incorrect.

• Buildings and roads can now have arbitrary user-defined fields on them.
• The largest zoom level at which the map is visible has been increased
• Buildings can be placed in groups, for which there is only a single connection decision (all or none of the buildings in a group must be connected). Groups can be inferred automatically during map creation from what road segment a building is connected to.
• Tiny holes in the network are now automatically closed during map creation.
• Model parameters can now be saved to and loaded from an excel spreadsheet.
• Underlying optimiser updated to a newer version.
• Project pages now show which problems have a solution. Running problems can be cancelled by clicking on the spinning thermos icon.
• The objective settings page now lets you stop the optimisation early if parameter fixing is not having a large effect on the result.
• To help find holes in the network, pipes which are connected to the network but go nowhere are now distinguished from those which are not connected at all.

• The combined demand profile for the supply model was computed incorrectly in some cases, making the total demand not match up with what the network model said. This is fixed.
• The network optimisation would sometimes fail if the a solution was feasible, but became infeasible when the diversity parameters were updated to reflect decisions in the first pass. This is now less likely to happen.
• The network optimisation model would not report correctly when a problem was infeasible.
• The candidate editor did not set connection costs when told to.
• It is no longer possible to tell the supply model that a technology is CHP but has no power efficiency; doing this this caused the supply model to fail.
• An incorrect bounds calculation for the maximum power flow on pipes to buildings with a connection count of more than 1 is fixed. This mistake made some solutions infeasible that shouldn't have been.
• The cooling peak demand was not assignable during map creation; it now is.

• You can now zoom the map in further.
• The info panel now has a row to tell you if you have edited a given candidate.
• The solution view in the map defaults to showing the pipe size.
• The splitters in the map part of the editor now have a button to minimize them.
• The changelog notification has been made more obnoxious.

• The network flow bounds calculation has been made faster.
• Network topology pre-processing no longer fails if the network contains a peripheral closed loop (a circular connected path that touches no buildings).
• Backward flow along connectors has been forbidden. This means that a network can no longer pass through a building.
• A bug has been removed which caused the model to ignore civil costs on some paths after they had been simplified to remove spurious junctions.

• When using the line drawing tool, if the new line split an existing line in two, each half of the split line would have the original length. The expected (and now correct) outcome is that the length of each half is recalculated. This fix does not apply retroactively to existing network problems.

• The variable capital cost for individual systems is now included. It was accidentally left out of the objective for the optimiser.