All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- Objective scaler for addressing problem ill-conditioning (#667)
- Workflow that ensures that CHANGELOG.md and the version number are updated (#711)
- Print GenX version at startup and export it to disk (#712)
- Option to output results with time series reconstructed for the entire year (#700)
- Default settings in multitage optimization (#703)
- Option to write results after each iteration of a multistage run (Myopic) (#704)
- Option to precompile
run_genx_case!
andcluster_inputs
to speedup TTFX of GenX (#718) - Co-location of electrolyzers to VRE_STOR and enabled regional hydrogen production limits (#678)
- Formatting and images in tutorials 3, 5, 6, and 8 to address issue #697 (#698)
devbranch
in the docsmake.jl
file to track the develop branch (#710)TDR
to loadFuels_data.csv
w/wo optionalNone
column (#720)- Include comprehensive list of all settings flags (that are currently in GenX) in the docs (#721)
- Allow
load_vre_stor_variability!
to load only wind or solar resources as part of the VRE_STOR module (#728) - Typo in HydrogenMinimumProduction setting key (#746)
- Default value of
Cap_Size
attribute for thermal generators (#747) - Text to doc pages for retrofits (#730)
- Generalized the cHourlyMatching constraint in electrolyzer.jl and implemented it similarly to the other policies, making it easier for developers to add further contributions to the constraint (#725).
- Feature CO2 and fuel module (#536) Adds a fuel module which enables modeling of fuel usage via (1) a constant heat rate and (2) piecewise-linear approximation of heat rate curves. Adds a CO2 module that determines the CO2 emissions based on fuel consumption, CO2 capture fraction, and whether the feedstock is biomass.
- Enable thermal power plants to burn multiple fuels (#586)
- Feature electrolysis basic (#525) Adds hydrogen electrolyzer model which enables the addition of hydrogen electrolyzer demands along with optional clean supply constraints.
- Add ability of storage to contribute to capacity reserves (#475)
- Add Co-located VRE+Storage Module (#523)
- Add output for dual of capacity constraint (#473)
- Add PR template (#516)
- Validation ensures that resource flags (THERM, HYDRO, LDS etc) are self-consistent (#513).
- Maintenance formulation for thermal-commit plants (#556).
- Add new tests for GenX: three-zone, multi-stage, electrolyzer, VRE+storage, piecewise_fuel+CO2, and TDR (#563 and #578).
- Added a DC OPF method (#543) to calculate power flows across all lines
- Added write_operating_reserve_price_revenue.jl to compute annual operating reserve and regulation revenue. Added the operating reserve and regulation revenue to net revenue (PR # 611)
- Add functions to compute conflicting constraints when model is infeasible if supported by the solver (#624).
- New settings parameter, VirtualChargeDischargeCost to test script and VREStor example case. The PR 608 attempts to introduce this parameter as cost of virtual charging and discharging to avoid unusual results (#608).
- New settings parameter, StorageVirtualDischarge, to turn storage virtual charging and discharging off if desired by the user (#638).
- Add module to retrofit existing resources with new technologies (#600).
- Formatted the code and added a format check to the CI pipeline (#673).
- Add check when capital recovery period is zero and investment costs are non-zero in multi-stage GenX (#666)
- Added condition number scaling added to objective function (#667)
- Added versioned doc-pages for v0.3.6 and v0.4.0
- Added a warning message in write_costs_multistage mentioning th approximate value of costs currently.
- Add constraint in mga to compute total capacity in each zone from a given technology type (#681)
- New settings parameter MGAAnnualGeneration to switch between different MGA formulations (#681)
- Add validation for
Can_Retire
column in multi-stage GenX since the current implementation does not allow a resource to switch from can_retire = 0 to can_retire = 1 between stages. (#683) - Add tutorials for running GenX (#637 and #685)
- Add writing of multistage stats during optimization with foresight (#687)
- Fix docstring in operational_reserves.jl (#690)
- Fix docstring in energy_share_requirement.jl (#692)
- Set MUST_RUN=1 for RealSystemExample/small_hydro plants (#517). Previously these plants had no resource flag set, and so they did not contribute to the power balance. As these plants are now useful, the objective in these cases is slightly lower.
- Assign correct investment cost per stage in objective function initialization for multi-stage runs (#530)
- Fix name of Fixed_OM_Cost_Charge_per_MWyr when reading from Generators_data in multistage code. (#533) Previously there was a typo in this in the multistage code that led to a silent bug, which affects outputs, for anyone running non-myopic multistage GenX with asymmetric storage.
- Fix computation of cumulative minimum capacity retirements in multistage GenX (#514)
- Fix access of eELOSSByZone expr before initialization (#541)
- Correctly write unmet reserves (in reg_dn.csv) (#575)
- Correctly scale total reserves column (in reg_dn.csv) (#594)
- Add validation for
Reg_Max
andRsv_Max
columns inGenerators_data.csv
whenMUST_RUN
is set to 1 (#576) - Fix scaling of transmission losses in write_transmission_losses.jl (#621)
- Fix cost assignment to virtual storage charge/discharge - issue #604 (#608)
- Fix modeling of hydro reservoir with long duration storage (#572).
- Fix update of starting transmission capacity in multistage GenX
- Fix write_status with UCommit = WriteShadowPrices = 1 (#645)
- Fixed outputting capital recovery cost to 0 if the remaining number of years is 0 (#666)
- Updated the docstring for the initialize_cost_to_go function and adjusted the formula for the discount factor to reflect the code implementation (#672).
- Fix write_multi_stage_cost.jl: add discount with OPEX multipliers to cUnmetPolicyPenalty (#679)
- Fix DF calculation in DDP to make it more generic for variable length stages (#680)
- Fix write_power_balance.jl: add additional two columns ("VRE_Storage_Discharge" and "VRE_Storage_Charge") for VRE_STOR
- Use add_to_expression! instead of the += and -= operators for memory performance improvements (#498).
- Generally, 'demand' is now used where (electrical) 'load' was used previously (#397).
Load_data.csv
is being renamed toDemand_data.csv
.Load_MW_z*
columns in that file are renamed toDemand_MW_z*
.- In
Reserves.csv
,Reg_Req_Percent_Load
andRsv_Req_Percent_Load
are renamed to..._Demand
. Load
andLoadWeight
keys in thetime_domain_reduction_settings.yml
file are renamed toDemand
,DemandWeight
.- The
New_Build
column inGenerators_data.csv
has been separated into two:New_Build
andCan_Retire
(#392). Values in each column are {0,1}. - Separate proprietary JuMP solvers from the GenX package. This allows users of Gurobi or CPLEX to use them without modifying the source code of the GenX package directly. This is a key step in publishing GenX as a proper Julia package. This does require change to the Run.jl files, to specify the solver. (#531)
- In the examples, change Reg_Max and Rsv_Max of any MUST_RUN generators to 0. This mitigates but does not fully fix (#576).
- Expressions of virtual charging and discharging costs in storage_all.jl and vre_stor.jl
- The input file
Generators_data.csv
has been split into different files, one for each type of generator. The new files are:Thermal.csv
,Hydro.csv
,Vre.csv
,Storage.csv
,Flex_demand.csv
,Must_run.csv
,Electrolyzer.csv
, andVre_stor.csv
. The examples have been updated, and new tests have been added to check the new data format (#612). - The settings parameter
Reserves
has been renamed toOperationalReserves
,Reserves.csv
toOperational_reserves.csv
, and the.jl
files contain the wordreserves
have been renamed tooperational_reserves
(#641). - New folder structure for a GenX case. The input files are now organized in the following folders:
settings
,policies
,resources
andsystem
. The examples and tests have been updated to reflect this change. - New folder structure implemented for
example_system
. This folder now consists of nine separate folders each pertaining to a different case study example, ranging from the ISONE three zones, with singlestage, multistage, electrolyzers, all the way to the 9 bus IEEE case for running DC-OPF. - Pruned HiGHS solver settings to the necessary minimum (#668)
- Changed deploydoc URL to GenX.jl (#662)
- The above
load
keys, which generally refer to electrical demand, are being deprecated. Users should update their case files. For now this is done in a backward-compatible manner, and @info reminders are written to the log to prompt the user to update. "Load" now typically refers only to the transferrence of data from files to memory, except for a few places such as the common term "value of lost load" which refers to non-served demand (#397). New_Build = -1
inGenerators_data.csv
: instead, useNew_Build = 0
andCan_Retire = 0
.- The matrix-style input of the grid for Network.csv is deprecated in favor a column-style input. Instead of columns z1, z2, ... with entries -1, 0, 1, use two columns: Start_Zone, End_Zone (#591).
- Order of slack policy constraint declarations (#464)
- Sign error in the Maximum Capacity Requirement slack constraint term (#461)
- Remove outdated HiGHS setting
simplex_dualise_strategy
(#489) - Fix bug in LDES outputs (#472)
- Updated README with new instructions for running GenX through Julia REPL terminal (#492)
- Fix factor of 0.5 when writing out transmission losses. (#480)
- Fix summation error when a set of hours is empty (in thermal_commit.jl).
- Fix access to eELOSSByZone expr before initialization (#541)
- Fix modeling of hydro reservoir with long duration storage (#572).
- Fix computation of cumulative minimum capacity retirements in multistage GenX (#514)
- Fix update of starting transmission capacity in multistage GenX
- Eliminate 'Axis contains one element' warning (seen when LDS is used) by combining two constraints (#496).
- Added ability to apply run_timedomainreduction to multistage problems (#443).
- Added a description of how to use time domain reduction (#426).
- Validation: against trying to perform time domain reduction (clustering) on data which has already been clustered.
- This changelog (#424).
- Added ability for CPLEX handle the Pre_Solve key (#467).
- run_timedomainreduction did not support multi-stage problems (#441).
- Not having a changelog (#423).
- The columns
Rep_Periods
,Timesteps_per_Rep_Period
, andSub_Weights
are now required inLoad_data.csv
for all cases (#426).
- The settings key
OperationsWrapping
. Its functionality has now been folded into theTimeDomainReduction
setting. Using the key now will print a gentle warning (#426).
- Validation of the time basis in
load_data.csv
(#413). - Arbitrary option keys can be passed to Solvers (#356).
- Validation for OperationWrapping and TimeDomainReduction settings (#337).
- Ability to use 'slack variables' to violate policy constraints---at a cost (#328, #435).
- Documented minimum up & down-time constraints (#324).
- Validation preventing two columns of input data with the same name (#309).
- Fuel type
None
is no longer need to be listed as a column infuels_data.csv
, e.g. for solar panels (#303). - Non-varying generators (e.g. thermal generators) no longer need to be listed in
generators_variability.csv
(#303). - Ability to load the transmission network representation from lists rather than a matrix (#292).
- Maximum Capacity Requirement policy constraint.
- New
run_genx_case!
function for use in scripts. - New
run_timedomainreduction!
function for pre-clustering a case. - Improved documentation.
- Corrected the interaction of Reserves and Regulation policies with ramp rates (#415).
- Removed the useless
MinCapTag
column from examples (#380). - Removed invalid
BarObjRng
key from gurobi settings (#374). - Default
crossover
orrun_crossover
settings (#363). - HYDRO resources now allow the period map to be loaded (#362).
- Numbering in documentation (#330).
- Correct scaling of emission outputs (#322).
- Add transmission losses to ESR constraints (#320).
- Author's name spelling in docs (#317).
- Unset executable bits on files (#297).
- Morris method example now runs.
- Various other settings issues with example cases.
- Simplified the
Simple_Test_Case
example (#414). - SmallNewEngland/Onezone example now uses linearized unit committment by default (#404).
- Removed the unused dependency BenchmarkTools (#381).
- The unmaintained MonteCarlo code (#357).
- License blocks from most file headers (#353).
- Extra
LDS
columns from several examples (#312). - SCIP from the Project and from documentation.