Capacity Expansion and Dispatch Learning Model v1.59.6

Start here: embedded Mini-grid case

Purpose. The embedded case is loaded on file open. It frames a two-node Mini-grid with demand at both nodes, supply and some storage at System Node, demand-side storage at Node 2, and a bidirectional shared-capacity Link 1 between the nodes. Use Load example setup only when you want to reset the working case back to the embedded baseline.

Recommended workflow

Dashboard: review economic horizon, cost of capital, Value of Lost Load, project notes and the node-link diagram.
Grid Load Builder: review load objects first. Demand is assigned to nodes through load objects, reference demand and PLEXOS-World profiles, not through the Network tab.
Resource profiles: confirm the solar and wind traces, location assumptions and loaded profile library.
Network: review nodes, label-only node roles, and transmission assets. Bidirectional links are linear shared-capacity transport links.
Technology Candidates: review supply, storage, costs, build limits, existing capacity and node assignment.
Settings: review pre-solve diagnostics, reliability constraints, storage settings and non-commercial guide weights.
Run optimisation: read Decision Summary first, then build, cost, network/storage diagnostics, stress, dispatch, detailed tables and finance.
Compare and export: use Fixed system comparator, flat-world sensitivity, stress tests and Scenario comparison before exporting dispatch or finance CSVs.

Embedded case framing

Item	Framing
Demand	876,000 MWh/y total, 100 MW average. System Node has 613,200 MWh/y, or 70 MW average, from the 60 MW block load and 10 MW commercial load. Node 2 has 262,800 MWh/y, or 30 MW average, from the 20 MW industrial load plus 10 MW average PLEXOS-World Japan demand.
Supply	Solar, wind, CCGT, OCGT and pumped hydro candidates are assigned to System Node. The BESS candidate is assigned to Node 2.
Transmission	Link 1 is bidirectional shared-capacity transport between System Node and Node 2, with 10 MW existing capacity, 250 MW maximum build and 1% loss.
Counterfactual	Flat-world sensitivity solves the same case as a single unconstrained hub. Treat it as a congestion-cost comparator, not as a physical network design.
Boundary	Linear transport screening model. No unit commitment, binary build decisions, complementarity, alternating-current power flow, security-constrained dispatch, market-price formation or nodal pricing.

Where to look after solve

Decision Summary gives the first-pass management readout.
Network & storage identifies link binding, losses, storage cycling and location signals.
Hourly dispatch explorer checks operational behaviour for the selected result, including flat-world and stress cases.
Detailed tables, audit and exports provides the dispatch CSV with object-level columns where solver detail is available.
Finance detail provides the system-wide profit and loss and discounted cash-flow export. It does not change the optimisation.

Model horizon, hours

Model start date

Economic Investment Horizon, years

Cost of Capital / Target IRR, real %

Tax rate, %

Currency

Carbon price, $/tCO2

Unserved energy penalty, $/MWh

Save, restore and share setup

Save the complete working setup as a portable JSON file. This includes system settings, candidate assets, loaded demand and renewable profiles, load-builder components, map/resource assumptions and optional solved results. Another user can open the same HTML file, load this JSON, and rerun or inspect the case.

Project name

Include solved results in standard saveInclude latest optimal and fixed system comparator results

Project notes

Load saved setup JSON

No saved setup loaded in this session.

Ready. Embedded Mini-grid two-node case is framed. Review demand, network and candidate assumptions, then run optimisation.

Dashboard abstraction: node-link diagramConfigured system

Conceptual nodal view. In one-node mode the existing hub-and-spoke view is retained. In multi-node mode the diagram uses node cards and directed or bidirectional link arrows, with technology candidates and storage grouped inside their assigned node. Drag node cards to arrange the diagram for readability. This is still a linear transport abstraction, not an alternating-current power-flow model.

Hover or click a hub-and-spoke box to view configured limits, optimised installed capacity and fixed system comparator framing.

This is a screening and teaching model. It excludes unit commitment, alternating-current power flow, security constraints and integer build blocks.

Publication, data provenance and license

Software license: GNU General Public License v3.0 or later (GPL-3.0-or-later). Users may use, share and modify the tool. If they distribute modified versions, the modified source should remain available under the same license.

PLEXOS-World demand data attribution: the embedded demand library is a compressed reconstruction derived from PLEXOS-World 2015, Harvard Dataverse, DOI 10.7910/DVN/CBYXBY. The specific demand-file reference used during development was doi:10.7910/DVN/CBYXBY/JCQPTE.

Recommended citation: Brinkerink, M., and Deane, P. (2020). PLEXOS-World 2015. Harvard Dataverse. Also cite the related article, Building and Calibrating a Country-Level Detailed Global Electricity Model Based on Public Data, when using the embedded demand profiles in published or shared analysis.

Netlify note: publish the generic app only. Do not publish saved setup JSON files containing confidential locations, demand traces, costs, capacities or commercial assumptions.

Resource coordinate frame

Use this section to pick a location and generate one-year solar and wind CSV download links from Renewables.ninja. Click or drag the map marker to set the target coordinate, download the PV and wind CSV files, then upload those files below. This section is for renewable resource data only.

Dataset

Latitude

Longitude

Location context map

Map marker sets the latitude and longitude used for solar and wind CSV downloads.

Reference Calendar Year

Download period

Renewables.ninja point downloads are framed as a single calendar-year request. Internal date fields are generated from the selected Reference Calendar Year.

Solar assumptions

PV capacity basis

System loss

Tracking

Tilt, degrees

Azimuth, degrees

Wind assumptions

Hub height, m

Turbine

Dropdown uses the built-in Renewables.ninja turbine preset list. Saved custom turbine values are preserved if loaded.

Recommended workflow: choose the location and assumptions above, then use the CSV buttons below. If the browser opens the CSV instead of downloading it, save the page as a .csv file and upload it here.

Renewables.ninja URL preview

No Renewables.ninja resource profiles loaded yet.

Resource CSV uploads

Upload solar and wind capacity-factor files. Files can be raw capacity factors from 0 to 1, percentages from 0 to 100, or native Renewables.ninja CSV exports. For Renewables.ninja files, the parser uses the electricity column, not wind_speed, irradiance or temperature.

Solar capacity factor or Renewables.ninja PV CSV Wind capacity factor or Renewables.ninja wind CSV

Loaded renewable profile library

Upload one or more PV and wind files. Each file is stored as a named profile that can be assigned to a specific solar or wind candidate under Technology Candidates. The most recently loaded profile also becomes the current default profile.

Class	Name	Rows	Average CF	Max CF	Source

Resource profile preview

Chart preview shows the first 168 hours of solar and wind capacity factors.

Reference demand CSV upload

Upload up to 8,760 hourly MW values as a reference grid-load profile. This can be previewed, applied directly to the model, or combined with a selected PLEXOS-World demand block and/or enabled Load Builder blocks through Demand Stack below.

Reference demand CSV, MW

No reference demand CSV loaded. The embedded case uses the active demand profile created from the Grid Load Builder components.

PLEXOS-World demand profile library

Use this to load a country or regional proxy demand shape from the embedded PLEXOS-World 2015 demand file, convert it from UTC into a local-year profile, then rescale it to the project size. The working convention is that the final model profile starts at 1 Jan 00:00 local time, aligned with the local-time treatment of Renewables.ninja exports.

Search PLEXOS region or country

Electricity Maps zone reference

Available PLEXOS region or country

Scaling method

Scaling value

Loaded profile rows

PLEXOS time basis

Offset source

Local UTC offset, hours

Time alignment not yet calculated.

Optional: replace embedded library with a downloaded PLEXOS file

PLEXOS-World demand file

Dataverse source

Embedded compressed PLEXOS-World demand library is ready. Search by country, region or PLEXOS code, then scale and apply a selected demand shape.

Load builder: block and seasonal loads

Build a local demand shape by combining load components. Each component has a 24-hour archetype and a 12-month seasonal scalar. The 24-hour shape is normalised to daily average 1.00, so the component's average MW remains intuitive.

Use	Name	Archetype	Average MW	Peak MW	Annual MWh

Selected component

Apply archetype preset

Normalise monthly scalars to an annual weighted average of 1.00. This keeps each component's average MW unchanged while still allowing months like Jan-Mar at 2.00x and Apr-Jun at 0.70x.

24-hour shape grid for selected component

Values are relative hourly scalars. They are normalised internally to daily average 1.00.

12-month seasonal scalar grid for selected component

Enter monthly multipliers. Example: Jan-Mar at 2.00, Apr-Jun at 0.70, other months at 1.00.

Load builder not initialised.

Demand stack: combine reference, PLEXOS and load-builder demand

Use this to combine an uploaded reference demand profile, a selected PLEXOS proxy profile and/or project-specific load-builder blocks. The combined profile starts at 1 Jan 00:00 local time.

Use uploaded reference demand
The loaded CSV profile is added to the framed demand stack. Use selected PLEXOS profile
Scaled using the PLEXOS controls above. Add enabled load-builder components
All enabled components in the load builder are added to the framed profile.

Demand stack not calculated yet.

Operational rule: PLEXOS-World can provide the base demand shape and the Load Builder can add project-specific loads. The scale remains a modelling assumption unless the user enters a peak, average or annual demand. The default applies a circular UTC-to-local-year shift so hour 0 means 1 Jan 00:00 local time. Electricity Maps coverage is used for zone-context alignment, not as a demand feed in this browser-only prototype.

Network

Network topology. Grid Load Builder defines demand. Network defines nodes, descriptive node roles, diagram positions and transmission assets. Enabled transmission assets are linear transport links with capacity, losses and costs.

Nodes

Assigned demand

0 MWh/y

Transmission assets

Nodes / regions

Default one-node mode uses System Node. Add nodes to activate the multi-node transport formulation. Demand is assigned in Grid Load Builder and audited read-only here by node.

Node name	Node role / label only	Annual demand	Peak MW	Average MW	Share	Diagram X	Diagram Y	Notes

Demand is assigned in Grid Load Builder. The Network tab audits assigned demand by node.

Transmission assets active in multi-node mode

Directed transfer limits, losses and costs become active when two nodes are framed. Links can be directed or bidirectional. Bidirectional convenience mode internally creates forward and reverse non-negative flow variables with shared capacity and shared build cost.

Enabled	Asset name	From node	To node	Existing MW	Max build MW	Loss %	Notes

Transmission is active in two-to-four-node mode. Directed links use one non-negative flow direction. Bidirectional links use a shared capacity where forward flow plus reverse flow cannot exceed total transfer capacity in an hour. In one-node mode this table remains saved metadata and the original single-node solve path is preserved.

Transmission links are linear transport constraints. The model does not perform alternating-current power flow, security-constrained dispatch, voltage or stability modelling, or nodal pricing.

Candidate assets

Add multiple candidates within each technology class. Solar and wind candidates sit under Renewables and can be assigned different uploaded resource profiles. The optimiser can build each candidate independently, subject to minimum and maximum build limits.

Capital cost is annualised against the Dashboard Economic Investment Horizon. If Tech. Life is longer than the economic horizon, capital is recovered within the economic horizon. If Tech. Life is shorter, the model adds a discounted full or partial refurbishment provision before the end of the economic horizon.

Technology Class: Renewables profile selectable by candidate

Solar candidates

Solar candidate fields are split across technical/resource settings and economics/build limits.

Wind candidates

Wind candidate fields are split across technical/resource settings and economics/build limits.

Technology Class: Firm Generators

Firm generator candidate fields are split across operating parameters and economics/build limits.

Technology Class: Energy Storage Systems

Energy storage candidate fields are split across technical settings and economics/build limits.

External Grid Interface

Interface import price, $/MWh

Interface emissions, tCO2/MWh

Max interface import, MW

Candidate detail is shown in the capacity table after optimisation. For firm generators, VOM $/MWh is calculated as fuel cost $/GJ x (3.6 / thermal efficiency fraction) + non-fuel VOM $/MWh. Loss is entered in percentage points: 0.12 means 0.12%, not 12%. Negative values are allowed as a modelling shorthand for hub-and-spoke representations where a generator is credited with a negative marginal loss factor at the supply-demand node. Parasitic MW/MW adds a constant load for every MW of installed capacity. Min build MW is a lower bound on new capacity built by the optimiser, not existing capacity. Aggregate charts still group solar, wind, firm and energy storage for readability. Candidate-level profile choice, build limits and capacity detail are retained in the capacity table and setup export.

Fixed system comparator

Use this to test a specified fixed system configuration. The model will run dispatch only: no new capacity is added. The installed capacity fields below are generated from the candidates framed in Technology Candidates, so the user can nominate exactly which solar, wind, firm generation and Energy Storage System candidate is installed.

Candidate-level installed capacityNo fixed system comparator capacities framed

These fields are dispatch inputs, not expansion limits. Candidate max-build values from Technology Candidates are shown for context, but the fixed system comparator can deliberately test a fixed configuration outside those bounds.

Class	Candidate	Technology frame	User installed MW	User installed MWh	Dispatch interpretation

Comparison basis: the optimiser still chooses capacity on an incremental basis, but reported system cost includes annualised fixed cost for total installed capacity. This makes an optimal greenfield solution and a candidate-level user-defined greenfield solution comparable.

Comparison summary

Documentation and model boundary

Current scope. Single-file browser-only linear capacity expansion and hourly dispatch screening tool. Demand is defined through load objects and demand profiles, Network defines nodes and transport links, Technology Candidates define supply and storage, and Results explain the decision case.

Workflow map

Frame demand: use Grid Load Builder load objects, reference demand and/or PLEXOS-World demand. Each demand source is assigned to a node.
Frame topology: use Network for node names, label-only node roles, diagram positions and transmission assets.
Frame supply: use Technology Candidates for renewables, firm generation, storage and the External Grid Interface.
Run and diagnose: use pre-solve diagnostics before solving, then use Decision Summary, network/storage diagnostics and dispatch charts after solving.
Compare cases: use stress tests, Fixed system comparator dispatch, flat-world sensitivity and Scenario comparison.
Export evidence: use dispatch CSV for hourly audit and finance CSV for discounted cash-flow review.

Load objects and demand framing

Load objects are first-class demand components with a name, average MW, 24-hour shape, 12-month scalar, enabled flag and node assignment.
The Network tab audits demand by node but does not allocate demand. This avoids two competing sources of demand truth.
Reference demand CSV and PLEXOS-World demand can be assigned to nodes and combined with load-builder demand in Demand Stack.
The pre-optimisation demand chart shows all-system demand and node-level demand so the user can check the load frame before solving.

Network and bidirectional links

Network links are linear transport constraints with existing capacity, optional build capacity, build cost, fixed operating cost, variable transfer cost and losses.
Directed links use one non-negative flow variable from origin to destination.
Bidirectional links use forward flow >= 0, reverse flow >= 0, and forward flow plus reverse flow <= total transfer capacity. Build cost is shared once across the bidirectional asset.
Bidirectional convenience mode remains linear. It does not model alternating-current power flow, loop flow, voltage, stability or security constraints.

Scenario comparison, saved results and restore

Scenario comparison stores compact snapshots for base, stress, fixed system comparator and flat-world cases. It is intended for screening-level comparison, not as a full scenario database.
Save setup JSON stores the working setup. Save setup + results JSON also stores solved result data where available.
Saved results can restore aggregate and object-level series where the saved package contains them. If a saved result lacks solver-column detail, rerun the case for the most complete hourly audit.
Restored results should be treated as evidence attached to the saved setup. If inputs are changed after load, rerun before relying on charts or exports.

Storage validity, guide scores and dispatch polish

Storage is linear and may show artefacts such as simultaneous charge/discharge or counter-cycling in degenerate cases.
Storage anti-cycling cost, where enabled, is part of the optimisation. It should be kept small and interpreted as a numerical hygiene measure.
Post-solve guide scores for storage throughput, counter-cycling, transmission circulation, dumped energy and unnecessary imports are diagnostic only. They do not enter annual cost, objective cost, LCOE or finance outputs.
Dispatch polish fixes solved capacities and can reduce visual artefacts. It is still a linear post-processing step and does not convert the model into unit commitment.

Flat-world sensitivity

Flat-world sensitivity solves the same framed demand and supply candidates as a single unconstrained hub.
Use it to quantify the commercial cost of transmission bottlenecks and nodal placement constraints.
It is not a valid physical network dispatch because it ignores transfer limits between nodes. Treat it as a comparator, not a build plan.
Flat-world hourly results can be plotted in the dispatch explorer when the sensitivity has been solved or restored from a saved results file.

Dispatch and finance exports

Dispatch CSV exports hourly demand, generation, storage, unserved energy, dumping, node and link series. Candidate-level and link-level columns are included where solver detail is available.
Finance CSV exports the post-processing discounted cash flow and system-wide profit and loss for the currently plotted result.
Finance outputs do not feed back into the optimiser. They are a reporting layer over the solved technical system.
Conditional LCOE is annual system cost divided by served demand. It is not a tariff, pool price, contract price or investment approval metric.

Formulation boundary

Linear optimisation only.
Transport bottlenecks, link losses and link build costs are represented. Alternating-current power flow and security-constrained dispatch are not.
No unit commitment, binary build decisions, complementarity constraints, minimum stable generation, start costs, ramping constraints, reserve co-optimisation, voltage/stability modelling, market-price formation or nodal pricing.
Reliability is represented through unserved energy and user-defined constraints, not probabilistic outage modelling.
Outputs are screening-grade decision evidence, not a bankable design, grid connection study or investment approval.

Licence, attribution and deployment hygiene

Software licence: GPL-3.0-or-later. Embedded demand-profile reconstruction is derived from PLEXOS-World 2015, Brinkerink and Deane (2020), Harvard Dataverse, DOI 10.7910/DVN/CBYXBY. Third-party services and libraries retain their own licences and terms.

Publish the generic app only. Do not publish saved setup JSON files containing confidential locations, demand traces, commercial costs, capacities or internal assumptions.

Status

Not run

Annual cost

Conditional LCOE

Emissions

Unserved energy

Dumped energy

Decision Summary

Decision framing converts the solved optimisation into plain-language conditions for success. It is generated from the current solved case and should be read before the detailed metrics.

Status

Not run

Run optimisation to generate the decision summary.

Binding and near-binding constraints

This panel explains why the solution looks the way it does using simple constraint diagnostics, not dual values.

Constraint area	Indicator	Extent	Interpretation
Not run	-	-	Run optimisation to populate this panel.

Commercial reality

Reliability and firming are economic choices. This panel translates the solved technical system into board-level commercial signals.

Not run

Run optimisation to populate this panel.

Scenario comparison and decision dashboard

Compares solved snapshots for the optimised base, fixed system comparator, stress cases and flat-world sensitivity. The first snapshot remains the base case for deltas. Snapshots store summary metrics only, not full dispatch arrays.

Status

Not run

Run optimisation and add solved results to compare cases.

No scenario snapshots saved yet.

Scenario	Type	Annual cost	LCOE	Unserved	Decision signal
No scenario snapshots saved yet. Run a case, then add the plotted or base result.

Stress test layer

Stress tests apply selected shocks to the current setup and report deltas versus the solved base case. They do not overwrite the base solution.

Demand +10%
Uniform 10% load uplift. Renewable output -10%
Capacity-factor haircut as a proxy for adverse weather or correlation. Storage RTE -5 pp
Round-trip efficiency haircut. Firm availability -10 pp
Forced-outage equivalent on firm candidates. Carbon price +100
Adds 100 currency units per tCO2.

Stress per-test solver time limit, seconds

Stress wall-clock timeout, seconds

Stress tests run full-year reruns one at a time. If a stress does not return within the wall-clock timeout, that row is marked as timed out and the next selected stress continues. Default and maximum wall-clock runtime is 20 minutes per rerun. Reduce candidate count or horizon if a case still times out.

Run the base optimisation first. Stress results do not overwrite the base solution.

Stress	Status	Plot	Cost delta	Conditional LCOE delta	Unserved delta	Emissions delta	Interpretation

Model boundary and limitations

This visible boundary is intentional. It reduces misuse risk and keeps the tool honest.

Transport bottlenecks, link losses and link build costs are represented. Voltage limits, inertia, frequency security and alternating-current power flow are not.
No unit commitment, minimum stable generation, starts, ramping limits, binary build decisions or complementarity constraints.
No market price formation, contract revenue, merchant risk, revenue sufficiency or nodal pricing.
No probabilistic outage modelling, extreme-weather tail risk or security-constrained dispatch.
Storage polish and non-commercial guide scores are diagnostics. They do not turn the model into a physical dispatch engine.
Outputs are screening-grade evidence, not a bankable system design or investment approval.

Installed capacity by technology

Annual energy by function

System cost bridge

This is a cost bridge, not a tariff, market price or investment answer. The conditional LCOE is annual all-in system cost divided by annual served demand for this feasible modelled system. It excludes risk, optionality, network security and revenue sufficiency.

Run the optimiser to populate the cost bridge. Denominator is served demand in MWh per year, not gross generation or Energy Storage System discharge. Treat conditional LCOE as one diagnostic, not the decision.

Cost by system function

Detailed cost components

Node operating audit linked to supply-demand stack

This audit uses the selected plotted result and the same time slicer as the hourly dispatch explorer. The Supply-demand stack remains the primary node chart; this panel quantifies net imports, gross exports, storage charge/discharge, dumping, unserved energy and balance residual for the selected node. It is an operational diagnostic over the solved transport model, not a power-flow study.

Node to review

Chart basis

Follow the Supply-demand stack node filter when it is set to a specific node.

Use the main supply-demand stack for visual shape. Use this audit view for imports, exports, storage sinks, dumping, unserved energy and node balance residual.

Run or plot a solved result to populate node-level operating charts.

Selected node operating audit

Metric	Selected range	Unit	Interpretation

Hourly dispatch explorer

Use the day-time slicer, or the persistent bottom slicer, to inspect full-day or custom hourly slices of the selected solved result. Demand-framing traces can show the full selected period, while dense operational charts are capped to a 30-day rendered window for browser performance. Chart X axes use timestamps from the model start date. Energy Storage System charging is shown below zero. Dumping means renewable energy available in that hour that is neither served nor stored.

Hourly balance: demand equals dispatched supply plus Energy Storage System discharge plus External Grid Interface plus unserved energy minus Energy Storage System charging. Renewable availability is also balanced: available solar equals dispatched solar plus dumped solar, and available wind equals dispatched wind plus dumped wind.

Demand being framed before optimisation

This chart is available before optimisation. It shows demand from load objects, reference CSV and PLEXOS-World sources against active model demand. Node demand is summarised below.

No demand profile has been framed yet.

Result set to plot

Plotted result

This changes only the operational charts below. Stress-test rows become available here after successful stress solves. Selecting Fixed system comparator will solve fixed system comparator dispatch first if needed, then plot it. KPI tiles and the main capacity and energy tables continue to show the optimised base system.

Day-time slicerSelected period: full year

Start day

Day 1Day 183Day 365

Days to show

Slice mode

Start hour within start day

Window length, hours

Use full selected days for normal review. Switch to custom hourly window to start within a day and inspect a precise number of hours, for example 6:00 for 18 hours. All charts, legends, tooltips and the persistent bottom slicer follow the selected result and window.

Start day

Days to show

Quick view

Selected hours

Stack order preset

Custom stack order

Allowed keys: solar, wind, dis, firm, grid, unserved. In node-filtered charts, delivered imports follow the grid position. This changes only the visual stacking order, not the optimisation result.

Supply-demand stack chart

Stacked supply shows how solar, wind, firm generation, Energy Storage System discharge, External Grid Interface imports and any unserved energy meet node demand over the selected period. Energy Storage System charging is shown below zero because it is load on the system.

The chart legend is generated from active candidate-level series in the selected time window. In multi-node cases, the node filter can show the system total or a selected node. Drag the legend box to reposition it.

Dumped power and Energy Storage System charging

Positive values show dumped renewable power. Negative values show Energy Storage System charging, which absorbs surplus before dumping if it is economic and capacity is available.

Energy Storage System state of charge

Hourly marginal and average cash cost

Shows the selected plotted result: optimised system, fixed system comparator, plotted stress case or flat-world sensitivity. Marginal cost is a short-run cash-cost proxy from the highest-cost dispatched block in each hour, with unserved energy setting the Value of Lost Load when present. Average cash cost is hourly variable, import, carbon, unserved and dumping cash cost divided by served node demand. It is not a market price and it excludes capacity recovery. Summary figures below show full-year and selected-range averages. Click or tap a point on this chart to set the selected hour used by the merit-order curve and SRMC demand-crossing stack.

Merit-order hour

Selected timestamp

The merit order is a cash-cost stack for the selected hour. It uses available candidate capacity in that hour, not a network dispatch curve. Storage is shown at its direct variable cost only; its opportunity value is not calculated without solver dual prices.

Merit order cost curve for selected hour

Run or plot a solved result to populate the merit-order chart.

Export checks

Checks the Dispatch CSV header for the currently plotted solved result. It confirms that demand, node, link-flow, storage, load-object, balance-residual and finance-export columns are available where expected, then builds a column dictionary from the observed export header.

Status

Not run

Run optimisation and plot a result, then review the export schema.

Run optimisation and plot a result, then review the dispatch export columns.

Area	Check	Status	Detail	Action
Export checks	Not run	Watch	Use the review button after a solved result is plotted.	Then export Dispatch CSV and Finance CSV.

Dispatch column dictionary preview

Preview shows the first dictionary rows. The download includes observed columns from the current plotted result where available, plus expected schema patterns.

Column or pattern	Area	Requirement	Definition	Interpretation
Not run	-	Watch	Run the column review.	The full dictionary is produced by the download button.

Detailed tables, audit and exports

Capacity table

Energy table

Energy balance audit

Optimal versus user system comparison

Finance detail: discounted cash flow and system-wide P&L

Post-processing finance view for the currently plotted result. The finance LCOE is solved so constant annual revenue from served energy produces a project IRR equal to the Cost of Capital / Target IRR, after fixed costs, variable costs, tax cash expense and straight-line depreciation over the Economic Investment Horizon. It does not change the optimisation.

System-wide P&L

Line item	Value	Unit	Interpretation

Annual discounted cash flow

Year	Revenue	Fixed cost	Variable cost	Depreciation	Tax	Net cash flow	Discounted cash flow	Cumulative discounted

Persistent day-time slicerSelected period: full year

Dynamic time slicerSelected period: full yearDrag the band or handles, like a pivot-chart timeline.

JanMarMayJulSepNovDec

Result set to plot

Slice mode

Start day

Days

Start hour

Window hours

Quick view

Selected hours

Start here: embedded Mini-grid case

Resource coordinate frame

Solar assumptions

Wind assumptions

Resource CSV uploads

Loaded renewable profile library

Reference demand CSV upload

PLEXOS-World demand profile library

Load builder: block and seasonal loads

Demand stack: combine reference, PLEXOS and load-builder demand

Network

Nodes / regions

Transmission assets active in multi-node mode

Candidate assets

Technology Class: Renewables profile selectable by candidate

Solar candidates

Wind candidates

Technology Class: Firm Generators

Technology Class: Energy Storage Systems

External Grid Interface

Fixed system comparator

Comparison summary

Controlled constraints

Optimisation settings

Documentation and model boundary

Decision Summary

Binding and near-binding constraints

Commercial reality

Scenario comparison and decision dashboard

Stress test layer

Installed capacity by technology

Annual energy by function

System cost bridge

Cost by system function

Detailed cost components

Selected node operating audit

Hourly dispatch explorer

Demand being framed before optimisation

Supply-demand stack chart

Dumped power and Energy Storage System charging

Energy Storage System state of charge

Hourly marginal and average cash cost

Merit order cost curve for selected hour

Detailed tables, audit and exports

Capacity table

Energy table

Energy balance audit

Optimal versus user system comparison

Finance detail: discounted cash flow and system-wide P&L

System-wide P&L

Annual discounted cash flow