Code Reference

TSO Case Data

The TSO schema was based on the MATPOWER formats for the network and generator cost data, supplemented with TESP data. Code in fncsTSO.py reads this data from a JSON file.

Transmission System Operator (TSO)

http://example.com/root.json
type	object
properties
version	The Version Schema
	not used
	type	integer
	examples	2
	default	0
baseMVA	The Basemva Schema
	MVA base for impedances
	type	integer
	examples	100
	default	0
StartTime	The Starttime Schema
	Date and time corresponding to 0 seconds in FNCS
	type	string
	examples	2013-07-01 00:00:00
	pattern	^(.*)$
	default
Tmax	The Tmax Schema
	Number of seconds to simulate
	type	integer
	examples	86400
	default	0
Period	The Period Schema
	Optimal power flow (OPF) interval in seconds
	type	integer
	examples	300
	default	0
dt	The Dt Schema
	Regular power flow (PF) interval in seconds
	type	integer
	examples	60
	default	0
pf_dc	The Pf_dc Schema
	1 for DC PF, 0 for AC PF
	type	integer
	examples	1
	default	0
opf_dc	The Opf_dc Schema
	1 for DC OPF, 0 for AC OPF
	type	integer
	examples	1
	default	0
bus	The Bus Schema
	Bus data, including loads and voltage base
	type	array
	items	Bus Array
		type	array
		items
			Bus Number
			type	number
			examples	1
			Type (1=load,2=gen,3=swing)
			type	number
			enum	1, 2, 3
			examples	3
			Pd (load)
			type	number
			examples	15167.5
			Qd (load)
			type	number
			examples	3079.89
			Gs (shunt MW)
			type	number
			examples	0
			Bs (shunt MVA)
			type	number
			examples	5000
			Area
			type	number
			examples	1
			V magnitude (pu)
			type	number
			examples	1
			V angle (deg)
			type	number
			examples	0
			kV base
			type	number
			examples	345
			Zone
			type	number
			examples	1
			Vmax pu
			type	number
			examples	1.1
			Vmin pu
			type	number
			examples	0.9
gen	The Gen Schema
	Generator ratings
	type	array
	items	Generator Array
		type	array
		items
			Bus
			type	number
			examples	1
			Pg (MW)
			type	number
			examples	0
			Qg (MVAR)
			type	number
			examples	0
			Qmax (MVAR)
			type	number
			examples	6567
			Qmin (MVAR)
			type	number
			examples	-6567
			Vg (pu)
			type	number
			examples	1
			MVA base
			type	number
			examples	19978.8
			Status (1 in service)
			type	number
			enum	0, 1
			examples	1
			Pmax (MW)
			type	number
			examples	19978.8
			Pmin (MW)
			type	number
			examples	1998
			Pc1
			type	number
			examples	0
			Pc2
			type	number
			examples	0
			Qc1min
			type	number
			examples	0
			Qc1max
			type	number
			examples	0
			Qc2min
			type	number
			examples	0
			Qc2max
			type	number
			examples	0
			AGC ramp rate
			type	number
			examples	0
			10-min ramp rate
			type	number
			examples	0
			30-min ramp rate
			type	number
			examples	0
		Reactive ramp rate
		type	number
		examples	0
	Area participation factor
	type	number
	examples	0
branch	The Branch Schema
	Lines and transformers; pu impedance and ratings
	type	array
	items	Branch Array
		type	array
		items
			From Bus
			type	number
			examples	5
			To Bus
			type	number
			examples	6
			R (pu)
			type	number
			examples	0.004237
			X (pu)
			type	number
			examples	0.035898
			B (pu)
			type	number
			examples	2.48325
			Rating A, short term (MVA)
			type	number
			examples	2168
			Rating B, long term (MVA)
			type	number
			examples	2168
			Rating C, emergency (MVA)
			type	number
			examples	2168
			Tap Ratio for Xfmrs (From/To)
			type	number
			examples	0
			Shift Angle for Xfmrs (deg)
			type	number
			examples	0
			Status (1 in service)
			type	number
			enum	0, 1
			examples	1
			Min Angle Difference From-To (deg)
			type	number
			examples	-360
			Max Angle Difference From-To (deg)
			type	number
			examples	360
areas	The Areas Schema
	PF areas are not currently used in TESP
	type	array
gencost	The Gencost Schema
	Cost functions for generators and dispatchable loads
	type	array
	items	Generator Cost Array
		Indexing must match the Generators
		type	array
		items
			Flag - 2 for polynomial, 1 for piecewise linear
			type	number
			examples	2
			Startup cost
			type	number
			examples	0
			Shutdown cost
			type	number
			examples	0
			Number of coefficients
			type	number
			examples	3
			C2 coefficient
			type	number
			examples	0.005
			C1 coefficient
			type	number
			examples	40
			C0 coefficient
			type	number
			examples	0
FNCS	The FNCS Schema
	FNCS topics, scaling factors and initial conditions at selected buses
	type	array
	items	FNCS Bus Array
		type	array
		items
			bus ID
			type	number
			examples	1
			FNCS Topic
			type	string
			examples	SUBSTATION1
			GridLAB-D Scale Factor
			type	number
			examples	792
			Nominal P in MW
			type	number
			examples	15167.5
			Nominal Q in MVAR
			type	number
			examples	3079.89
			Scale factor for curve load
			type	number
			examples	0.5
			Skew for curve load in seconds
			type	number
			examples	1711
			Estimated P at time 0
			type	number
			examples	4788.99
			Estimated Q at time 0
			type	number
			examples	972.66
UnitsOut	The Units Out Schema
	Schedule of generators out of service
	type	array
	items	Generator Outage Array
		type	array
		items
			Index into Generators
			type	number
			examples	1
			Time outage starts in seconds
			type	number
			examples	108000
			Time outage ends in seconds
			type	number
			examples	154000
BranchesOut	The Branches Out Schema
	Schedules of branches out of service
	type	array
	items	Branch Outage Array
		type	array
		items
			Index into Branches
			type	number
			examples	2
			Time outage starts in seconds
			type	number
			examples	108000
			Time outage ends in seconds
			type	number
			examples	154000
swing_bus	The Swing_bus Schema
	Swing bus designation, depends on unit commitment
	type	integer
	examples	1
	default	0
definitions

src Directory Structure

This list shows directories and Python files under the tesp/src repository. On GitHub, each README contains a list of other files.

archive
- pypower; legacy files to patch PYPOWER; we have been able to incorporate these patches into the main PYPOWER distribution.
energyplus; C++ code to build a simple interface agent for EnergyPlus; this is part of the TESP distribution and used in the te30, sgip1 and energyplus examples.
gridlabd; legacy files for the house populations and feeder growth model; these features are mostly subsumed into tesp_support
jupyter; a prototype Jupyter notebook used for post-processing demonstrations and training
matpower
- ubuntu; legacy code that wraps MATPOWER for TESP, but only on Ubuntu. We now use PYPOWER. In 2017, the wrapping process was very difficult on Mac OS X, and unsuccessful on Windows using free compilers.
tesp_support; runs PYPOWER without FNCS
- setup.py; contains the version number and dependencies for tesp_support package
- tesp_support; Python code for agents, configuration and post-processing.
  - TMY2EPW.py; command-line script that converts a TMY2 file to the EnergyPlus EPW format.
  - TMY3toCSV.py; converts TMY3 weather data to CSV format for common use by agents.
  - __init__.py; boilerplate for a Python package
  - api.py; collects Python import statements needed to use public functions from Python code outside of this directory.
  - auction.py; supervises one double-auction and multiple HVAC agents for a feeder; communicates via FNCS with GridLAB-D and PYPOWER/AMES
  - feederGenerator.py; from a PNNL taxonomy feeder as the backbone, populates it with houses, solar PV, batteries and smart inverters
  - fncs.py; the Python interface to FNCS, which is a C/C++ shared object library, or dynamic link library (Windows)
  - tso_PYPOWER.py; manages PYPOWER solutions for the te30 and sgip1 examples, based on a 9-bus textbook model. Note that the ERCOT cases use custom local versions of this code instead.
  - glm_dict.py; parses the GridLAB-D input (GLM) file and produces metafile data in JSON format, describing the houses, meters, DER, capacitors and regulators
  - precool.py; manages a set of house thermostats for NIST TE Challenge 2. There is no communication with a market. If the house experiences an overvoltage, the thermostat is turned down and locked for 4 hours, unless the house temperature violates comfort limits.
  - prep_auction.py; configures the agent metadata (JSON) and GridLAB-D FNCS subscriptions/publications for the double-auction, double-ramp simulations
  - prep_precool.py; configures the agent metadata (JSON) and GridLAB-D FNCS subscriptions/publications for NIST TE Challenge 2 precooling
  - process_agents.py; makes tabular and plotted summaries of agent results
  - process_eplus.py; makes tabular and plotted summaries of EnergyPlus results
  - process_gld.py; makes tabular and plotted summaries of GridLAB-D results (substation power/losses, average and sample house temperatures, meter voltage min/max)
  - process_houses.py; plots the HVAC power and air temperature for all houses
  - process_inv.py; makes tabular and plotted summaries of results for NIST TE Challenge 2, including inverters, capacitor switching and tap changes
  - process_pypower.py; makes tabular and plotted summaries of PYPOWER results for the 9-bus model in te30 or sgip1
  - process_voltages.py; plots the minimum and maximum voltage for all houses
  - simple_auction.py; implements the double-auction agent and the Olympic Peninsula cooling agent, as separate Python classes, called by auction.py
  - tesp_case.py; supervises the assembly of a TESP case with one feeder, one EnergyPlus building and one PYPOWER model. Reads the JSON file from tesp_config.py
  - tesp_config.py; a GUI for creating the JSON file used to configure a TESP case
  - tesp_monitor.py; a GUI for launching a TESP simulation, monitoring its progress, and terminating it early if necessary
  - weatherAgent.py; publishes weather and forecasts based on a CSV file
  - README.md; this file
  - matpower; legacy code that configures and post-processes MATPOWER v5+ for TESP. We now use PYPOWER and AMES instead.
  - sgip1; custom code that plotted curves from different cases on the same graph. Used for a 2018 journal paper on TESP and the SGIP1 example.
  - valuation; custom code that post-processed SGIP1 outputs for the 2018 journal paper. May serve as an example, or use Jupyter notebooks instead.
- test; scripts that support testing the package; not automated.

Links to Dependencies

tesp_support package

TESP is the Transactive Energy Simulation Platform tesp_support contains the Python files that are part of TESP

Submodules

tesp_support.TMY2EPW module

Functions to convert Typical Meteorological Year data for EnergyPlus

The input must be in TMY2 format. If only TMY3 files are available, see the TMY3toTMY2_ansi.c program distributed with TESP under the support/weather/TMY2EPW/source_code directory.

Public Functions:

convert_tmy2_to_epw: Converts TMY2 file to EPW.

tesp_support.TMY2EPW.convert_tmy2_to_epw(fileroot)

Converts TMY2 to EPW

Reads fileroot.tmy2 and writes fileroot.epw

Parameters: fileroot (str) – The input path and base file name, without extension

tesp_support.TMY2EPW.removeZero(x): helper function to strip leading ‘0’ from a string

tesp_support.TMY3toCSV module

TMY3toCSV

Convert typical meteorological year version 3 (TMY3) data to comma separated values (CSV) for common use by TESP agents.

tesp_support.TMY3toCSV.readtmy3(filename=None, coerce_year=None, recolumn=True): Read a TMY3 file in to a pandas dataframe.

tesp_support.TMY3toCSV.weathercsv(tmyfile, outputfile, start_time, end_time, year_)

Converts TMY3 weather data to CSV in the requested time range

Parameters

tmyfile – string the input TMY3 data file
outputfile – string the output CSV data file
start_time – string the starting date and time of interest, like 2013-07-01 00:00:00
end_time – string the ending date and time of interest, like 2013-07-08 00:00:00
year – integer the year of interest

Returns

nothing

tesp_support.TMY3toCSV.weathercsv_cloudy_day(start_time, end_time, outputfile)

tesp_support.api module

tesp_support.case_merge module

Combines GridLAB-D and agent files to run a multi-feeder TESP simulation

Public Functions:

merge_glm: combines GridLAB-D input files
merge_glm_dict: combines GridLAB-D metadata files
merge_agent_dict: combines the substation agent configuration files
merge_substation_yaml: combines the substation agent FNCS publish/subscribe files
merge_fncs_config: combines GridLAB-D FNCS publish/subscribe files
merge_gld_msg: combines GridLAB-D HELICS publish/subscribe configurations
merge_substation_msg: combines the substation agent HELICS publish/subscribe configurations

tesp_support.case_merge.key_present(val, ary)

tesp_support.case_merge.merge_agent_dict(target, sources, xfmva)

Combines the substation agent configuration files into target/target.json. The source files must already exist.

Parameters

target (str) – the directory and root case name
sources (list) – list of feeder names in the target directory to merge

tesp_support.case_merge.merge_fncs_config(target, sources)

Combines GridLAB-D input files into target/target.txt. The source feeders must already exist.

Parameters

target (str) – the directory and root case name
sources (list) – list of feeder names in the target directory to merge

tesp_support.case_merge.merge_gld_msg(target, sources)

tesp_support.case_merge.merge_glm(target, sources, xfmva)

Combines GridLAB-D input files into target/target.glm. The source files must already exist.

Parameters

target (str) – the directory and root case name
sources (list) – list of feeder names in the target directory to merge

tesp_support.case_merge.merge_glm_dict(target, sources, xfmva)

Combines GridLAB-D metadata files into target/target.json. The source files must already exist.

Each constituent feeder has a new ID constructed from the NamePrefix + original base_feeder, then every child object on that feeder will have its feeder_id, originally network_node, changed to match the new one.

Parameters

target (str) – the directory and root case name
sources (list) – list of feeder names in the target directory to merge
xfmva (int) –

tesp_support.case_merge.merge_substation_msg(target, sources)

tesp_support.case_merge.merge_substation_yaml(target, sources)

Combines GridLAB-D input files into target/target.yaml. The source files must already exist.

Parameters

target (str) – the directory and root case name
sources (list) – list of feeder names in the target directory to merge

tesp_support.feederGenerator module

Replaces ZIP loads with houses, and optional storage and solar generation.

As this module populates the feeder backbone with houses and DER, it uses the Networkx package to perform graph-based capacity analysis, upgrading fuses, transformers and lines to serve the expected load. Transformers have a margin of 20% to avoid overloads, while fuses have a margin of 150% to avoid overloads. These can be changed by editing tables and variables in the source file.

There are two kinds of house populating methods implemented:

Feeders with Service Transformers

This case applies to the full PNNL taxonomy feeders. Do not specify the taxchoice argument to populate_feeder. Each service transformer receiving houses will have a short service drop and a small number of houses attached.
Feeders without Service Transformers

This applies to the reduced-order ERCOT feeders. To invoke this mode, specify the taxchoice argument to populate_feeder. Each primary load to receive houses will have a large service transformer, large service drop and large number of houses attached.

References

GridAPPS-D Feeder Models <https://github.com/GRIDAPPSD/Powergrid-Models>

Public Functions:

populate_feeder: processes one GridLAB-D input file

Todo

Verify the level zero mobile home thermal integrity properties; these were copied from the MATLAB feeder generator

tesp_support.feederGenerator.Find1PhaseXfmr(kva)

Select a standard 1-phase transformer size, with data

Standard sizes are 5, 10, 15, 25, 37.5, 50, 75, 100, 167, 250, 333 or 500 kVA

Parameters: kva (float) – the minimum transformer rating
Returns: the kva, %r, %x, %no-load loss, %magnetizing current
Return type: [float,float,float,float,float]

tesp_support.feederGenerator.Find1PhaseXfmrKva(kva)

Select a standard 1-phase transformer size, with some margin

Standard sizes are 5, 10, 15, 25, 37.5, 50, 75, 100, 167, 250, 333 or 500 kVA

Parameters: kva (float) – the minimum transformer rating
Returns: the kva size, or 0 if none found
Return type: float

tesp_support.feederGenerator.Find3PhaseXfmr(kva)

Select a standard 3-phase transformer size, with data

Standard sizes are 30, 45, 75, 112.5, 150, 225, 300, 500, 750, 1000, 1500, 2000, 2500, 3750, 5000, 7500 or 10000 kVA

Parameters: kva (float) – the minimum transformer rating
Returns: the kva, %r, %x, %no-load loss, %magnetizing current
Return type: [float,float,float,float,float]

tesp_support.feederGenerator.Find3PhaseXfmrKva(kva)

Select a standard 3-phase transformer size, with some margin

Standard sizes are 30, 45, 75, 112.5, 150, 225, 300, 500, 750, 1000, 1500, 2000, 2500, 3750, 5000, 7500 or 10000 kVA

Parameters: kva (float) – the minimum transformer rating
Returns: the kva size, or 0 if none found
Return type: float

tesp_support.feederGenerator.FindFuseLimit(amps)

Find a Fuse size that’s unlikely to melt during power flow

Will choose a fuse size of 40, 65, 100 or 200 Amps. If that’s not large enough, will choose a recloser size of 280, 400, 560, 630 or 800 Amps. If that’s not large enough, will choose a breaker size of 600 (skipped), 1200 or 2000 Amps. If that’s not large enough, will choose 999999.

Parameters: amps (float) – the maximum load current expected; some margin will be added
Returns: the GridLAB-D fuse size to insert
Return type: float

tesp_support.feederGenerator.ProcessTaxonomyFeeder(outname, rootname, vll, vln, avghouse, avgcommercial)

Parse and re-populate one backbone feeder, usually but not necessarily one of the PNNL taxonomy feeders

This function:

reads and parses the backbone model from rootname.glm

replaces loads with houses and DER

upgrades transformers and fuses as needed, based on a radial graph analysis

writes the repopulated feeder to outname.glm

Parameters

outname (str) – the output feeder model name
rootname (str) – the input (usually taxonomy) feeder model name
vll (float) – the feeder primary line-to-line voltage
vln (float) – the feeder primary line-to-neutral voltage
avghouse (float) – the average house load in kVA
avgcommercial (float) – the average commercial load in kVA, not used

tesp_support.feederGenerator.accumulate_load_kva(data)

Add up the total kva in a load-bearing object instance

Considers constant_power_A/B/C/1/2/12 and power_1/2/12 attributes

Parameters: data (dict) – dictionary of data for a selected GridLAB-D instance

tesp_support.feederGenerator.buildingTypeLabel(rgn, bldg, ti)

Formatted name of region, building type name and thermal integrity level

Parameters

rgn (int) – region number 1..5
bldg (int) – 0 for single-family, 1 for apartment, 2 for mobile home
ti (int) – thermal integrity level, 0..6 for single-family, only 0..2 valid for apartment or mobile home

tesp_support.feederGenerator.checkResidentialBuildingTable(): Verify that the regional building parameter histograms sum to one

tesp_support.feederGenerator.connect_ercot_commercial(op)

For the reduced-order ERCOT feeders, add a billing meter to the commercial load points, except small ZIPLOADs

Parameters: op (file) – an open GridLAB-D input file

tesp_support.feederGenerator.connect_ercot_houses(model, h, op, vln, vsec)

For the reduced-order ERCOT feeders, add houses and a large service transformer to the load points

Parameters

model (dict) – the parsed GridLAB-D model
h (dict) – the object ID hash
op (file) – an open GridLAB-D input file
vln (float) – the primary line-to-neutral voltage
vsec (float) – the secondary line-to-neutral voltage

tesp_support.feederGenerator.identify_ercot_houses(model, h, t, avgHouse, rgn)

For the reduced-order ERCOT feeders, scan each primary load to determine the number of houses it should have

Parameters

model (dict) – the parsed GridLAB-D model
h (dict) – the object ID hash
t (str) – the GridLAB-D class name to scan
avgHouse (float) – the average house load in kva
rgn (int) – the region number, 1..5

tesp_support.feederGenerator.identify_xfmr_houses(model, h, t, seg_loads, avgHouse, rgn)

For the full-order feeders, scan each service transformer to determine the number of houses it should have

Parameters

model (dict) – the parsed GridLAB-D model
h (dict) – the object ID hash
t (str) – the GridLAB-D class name to scan
seg_loads (dict) – dictionary of downstream load (kva) served by each GridLAB-D link
avgHouse (float) – the average house load in kva
rgn (int) – the region number, 1..5

tesp_support.feederGenerator.is_edge_class(s)

Identify switch, fuse, recloser, regulator, transformer, overhead_line, underground_line and triplex_line instances

Edge class is networkx terminology. In GridLAB-D, edge classes are called links.

Parameters: s (str) – the GridLAB-D class name
Returns: True if an edge class, False otherwise
Return type: Boolean

tesp_support.feederGenerator.is_node_class(s)

Identify node, load, meter, triplex_node or triplex_meter instances

Parameters: s (str) – the GridLAB-D class name
Returns: True if a node class, False otherwise
Return type: Boolean

tesp_support.feederGenerator.log_model(model, h)

Prints the whole parsed model for debugging

Parameters

model (dict) – parsed GridLAB-D model
h (dict) – object ID hash

tesp_support.feederGenerator.obj(parent, model, line, itr, oidh, octr)

Store an object in the model structure

Parameters

parent (str) – name of parent object (used for nested object defs)
model (dict) – dictionary model structure
line (str) – glm line containing the object definition
itr (iter) – iterator over the list of lines
oidh (dict) – hash of object id’s to object names
octr (int) – object counter

Returns

the current line and updated octr

Return type

str, int

tesp_support.feederGenerator.populate_all_feeders(): Wrapper function that batch processes all taxonomy feeders in the casefiles table (see source file)

tesp_support.feederGenerator.populate_feeder(configfile=None, config=None, taxconfig=None)

Wrapper function that processes one feeder. One or two keyword arguments must be supplied.

Parameters

configfile (str) – JSON file name for the feeder population data, mutually exclusive with config
config (dict) – dictionary of feeder population data already read in, mutually exclusive with configfile
taxconfig (dict) – dictionary of custom taxonomy data for ERCOT processing

tesp_support.feederGenerator.randomize_commercial_skew()

tesp_support.feederGenerator.randomize_residential_skew()

tesp_support.feederGenerator.randomize_skew(value, skew_max)

tesp_support.feederGenerator.replace_commercial_loads(model, h, t, avgBuilding)

For the full-order feeders, scan each load with load_class==C to determine the number of zones it should have

Parameters

model (dict) – the parsed GridLAB-D model
h (dict) – the object ID hash
t (str) – the GridLAB-D class name to scan
avgBuilding (float) – the average building in kva

tesp_support.feederGenerator.selectResidentialBuilding(rgnTable, prob)

Selects the building with region and probability

Parameters

rgnTable –
prob –

tesp_support.feederGenerator.selectSetpointBins(bldg, rand)

Randomly choose a histogram row from the cooling and heating setpoints

The random number for the heating and cooling set points row is generated internally.

Parameters

bldg (int) – 0 for single-family, 1 for apartment, 2 for mobile home
rand (float) – random number [0..1] for the cooling setpoint row

tesp_support.feederGenerator.selectThermalProperties(bldgIdx, tiIdx)

Retrieve the building thermal properties for a given type and integrity level

Parameters

bldgIdx (int) – 0 for single-family, 1 for apartment, 2 for mobile home
tiIdx (int) – 0..6 for single-family, 0..2 for apartment or mobile home

tesp_support.feederGenerator.union_of_phases(phs1, phs2)

Collect all phases on both sides of a connection

Parameters

phs1 (str) – first phasing
phs2 (str) – second phasing

Returns

union of phs1 and phs2

Return type

str

tesp_support.feederGenerator.write_commercial_loads(rgn, key, op)

Put commercial building zones and ZIP loads into the model

Parameters

rgn (int) – region 1..5 where the building is located
key (str) – GridLAB-D load name that is being replaced
op (file) – open file to write to

tesp_support.feederGenerator.write_config_class(model, h, t, op)

Write a GridLAB-D configuration (i.e. not a link or node) class

Parameters

model (dict) – the parsed GridLAB-D model
h (dict) – the object ID hash
t (str) – the GridLAB-D class
op (file) – an open GridLAB-D input file

tesp_support.feederGenerator.write_ercot_small_loads(basenode, op, vnom)

For the reduced-order ERCOT feeders, write loads that are too small for houses

Parameters

basenode (str) – the GridLAB-D node name
op (file) – an open GridLAB-D input file
vnom (float) – the primary line-to-neutral voltage

tesp_support.feederGenerator.write_houses(basenode, op, vnom, bIgnoreThermostatSchedule=True, bWriteService=True, bTriplex=True, setpoint_offset=1.0)

Put houses, along with solar panels and batteries, onto a node

Parameters

basenode (str) – GridLAB-D node name
op (file) – open file to write to
vnom (float) – nominal line-to-neutral voltage at basenode

tesp_support.feederGenerator.write_kersting_quadriplex(fp, kva)

Writes a quadriplex_line_configuration based on 1/0 AA example from Kersting’s book

The conductor capacity is 202 amps, so the number of triplex in parallel will be kva/sqrt(3)/0.208/202

tesp_support.feederGenerator.write_kersting_triplex(fp, kva)

Writes a triplex_line_configuration based on 1/0 AA example from Kersting’s book

The conductor capacity is 202 amps, so the number of triplex in parallel will be kva/0.12/202

tesp_support.feederGenerator.write_link_class(model, h, t, seg_loads, op, want_metrics=False)

Write a GridLAB-D link (i.e. edge) class

Parameters

model (dict) – the parsed GridLAB-D model
h (dict) – the object ID hash
t (str) – the GridLAB-D class
seg_loads (dict) – a dictionary of downstream loads for each link
op (file) – an open GridLAB-D input file

tesp_support.feederGenerator.write_local_triplex_configurations(op)

Write a 4/0 AA triplex configuration

Parameters: op (file) – an open GridLAB-D input file

tesp_support.feederGenerator.write_node_house_configs(fp, xfkva, xfkvll, xfkvln, phs, want_inverter=False)

Writes transformers, inverter settings for GridLAB-D houses at a primary load point.

An aggregated single-phase triplex or three-phase quadriplex line configuration is also written, based on estimating enough parallel 1/0 AA to supply xfkva load. This function should only be called once for each combination of xfkva and phs to use, and it should be called before write_node_houses.

Parameters

fp (file) – Previously opened text file for writing; the caller closes it.
xfkva (float) – the total transformer size to serve expected load; make this big enough to avoid overloads
xfkvll (float) – line-to-line voltage [kV] on the primary. The secondary voltage will be 208 three-phase
xfkvln (float) – line-to-neutral voltage [kV] on the primary. The secondary voltage will be 120/240 for split secondary
phs (str) – either ‘ABC’ for three-phase, or concatenation of ‘A’, ‘B’, and/or ‘C’ with ‘S’ for single-phase to triplex
want_inverter (boolean) – True to write the IEEE 1547-2018 smarter inverter function setpoints

tesp_support.feederGenerator.write_node_houses(fp, node, region, xfkva, phs, nh=None, loadkw=None, house_avg_kw=None, secondary_ft=None, storage_fraction=0.0, solar_fraction=0.0, electric_cooling_fraction=0.5, node_metrics_interval=None, random_seed=False)

Writes GridLAB-D houses to a primary load point.

One aggregate service transformer is included, plus an optional aggregate secondary service drop. Each house has a separate meter or triplex_meter, each with a common parent, either a node or triplex_node on either the transformer secondary, or the end of the service drop. The houses may be written per phase, i.e., unbalanced load, or as a balanced three-phase load. The houses should be #included into a master GridLAB-D file. Before using this function, call write_node_house_configs once, and only once, for each combination xfkva/phs that will be used.

Parameters

fp (file) – Previously opened text file for writing; the caller closes it.
node (str) – the GridLAB-D primary node name
region (int) – the taxonomy region for housing population, 1..6
xfkva (float) – the total transformer size to serve expected load; make this big enough to avoid overloads
phs (str) – ‘ABC’ for three-phase balanced distribution, ‘AS’, ‘BS’, or ‘CS’ for single-phase triplex
nh (int) – directly specify the number of houses; an alternative to loadkw and house_avg_kw
loadkw (float) – total load kW that the houses will represent; with house_avg_kw, an alternative to nh
house_avg_kw (float) – average house load in kW; with loadkw, an alternative to nh
secondary_ft (float) – if not None, the length of adequately sized secondary circuit from transformer to the meters
electric_cooling_fraction (float) – fraction of houses to have air conditioners
solar_fraction (float) – fraction of houses to have rooftop solar panels
storage_fraction (float) – fraction of houses with solar panels that also have residential storage systems
node_metrics_interval (int) – if not None, the metrics collection interval in seconds for houses, meters, solar and storage at this node
random_seed (boolean) – if True, reseed each function call. Default value False provides repeatability of output.

tesp_support.feederGenerator.write_one_commercial_zone(bldg, op)

Write one pre-configured commercial zone as a house

Parameters

bldg – dictionary of GridLAB-D house and zipload attributes
op (file) – open file to write to

tesp_support.feederGenerator.write_small_loads(basenode, op, vnom)

Write loads that are too small for a house, onto a node

Parameters

basenode (str) – GridLAB-D node name
op (file) – open file to write to
vnom (float) – nominal line-to-neutral voltage at basenode

tesp_support.feederGenerator.write_solar_inv_settings(op)

Writes volt-var and volt-watt settings for solar inverters

Parameters: op (file) – an open GridLAB-D input file

tesp_support.feederGenerator.write_substation(op, name, phs, vnom, vll)

Write the substation swing node, transformer, metrics collector and fncs_msg object

Parameters

op (file) – an open GridLAB-D input file
name (str) – node name of the primary (not transmission) substation bus
phs (str) – primary phasing in the substation
vnom (float) – not used
vll (float) – feeder primary line-to-line voltage

tesp_support.feederGenerator.write_tariff(op)

Writes tariff information to billing meters

Parameters: op (file) – an open GridLAB-D input file

tesp_support.feederGenerator.write_voltage_class(model, h, t, op, vprim, vll, secmtrnode)

Write GridLAB-D instances that have a primary nominal voltage, i.e., node, meter and load

Parameters

model (dict) – a parsed GridLAB-D model
h (dict) – the object ID hash
t (str) – the GridLAB-D class name to write
op (file) – an open GridLAB-D input file
vprim (float) – the primary nominal line-to-neutral voltage
vll (float) – the primary nominal line-to-line voltage
secmtrnode (dict) – key to [transfomer kva, phasing, nominal voltage] by secondary node name

tesp_support.feederGenerator.write_xfmr_config(key, phs, kvat, vnom, vsec, install_type, vprimll, vprimln, op)

Write a transformer_configuration

Parameters

key (str) – name of the configuration
phs (str) – primary phasing
kvat (float) – transformer rating in kVA
vnom (float) – primary voltage rating, not used any longer (see vprimll and vprimln)
vsec (float) – secondary voltage rating, should be line-to-neutral for single-phase or line-to-line for three-phase
install_type (str) – should be VAULT, PADMOUNT or POLETOP
vprimll (float) – primary line-to-line voltage, used for three-phase transformers
vprimln (float) – primary line-to-neutral voltage, used for single-phase transformers
op (file) – an open GridLAB-D input file

tesp_support.fncs module

Functions that provide access from Python to the FNCS library

Notes

Depending on the operating system, libfncs.dylib, libfncs.dll or libfncs.so must already be installed. Besides the defined Python wrapper functions, these pass-through library calls are always needed:

fncs.finalize: call after the simulation completes
fncs.time_request (long long): request the next time step; blocks execution of this process until FNCS grants the requested time. Then, the process should check for messages from FNCS.

These pass-through calls are also available, but not used in TESP:

fncs.route
fncs.update_time_delta
fncs.get_id
fncs.get_simulator_count
fncs.get_events_size
fncs.get_keys_size
fncs.die: stops FNCS and sends ‘die’ to other simulators

References

ctypes

FNCS

Examples

under tesp_support, see substation.py, precool.py and tso_PYPOWER_f.py
under examples, see loadshed/loadshed.py

tesp_support.fncs.agentGetEvents()

Retrieve FNCS agent messages

Returns: concatenation of agent messages
Return type: str

tesp_support.fncs.agentPublish(value)

Publish a value over FNCS, under the configured simulator name / agent name

Parameters: value (str) – value

tesp_support.fncs.agentRegister(config=None)

Initialize the FNCS configuration for the agent interface

Parameters: config (str) – a ZPL file. If None (default), provide YAML file in FNCS_CONFIG_FILE environment variable.

tesp_support.fncs.die(): Call FNCS die because of simulator error

tesp_support.fncs.finalize(): Call FNCS finalize to end connection with broker

tesp_support.fncs.get_event_at(i)

Retrieve FNCS message by index number after time_request returns

Returns: one decoded FNCS event
Return type: str

tesp_support.fncs.get_events()

Retrieve FNCS messages after time_request returns

Returns: tuple of decoded FNCS events
Return type: list

tesp_support.fncs.get_events_size(): Get the size of the event queue

tesp_support.fncs.get_id(): Find the FNCS ID

tesp_support.fncs.get_key_at(i)

Get the topic by index number

Parameters: i (int) – the index number
Returns: decoded topic name
Return type: str

tesp_support.fncs.get_keys()

Find the list of topics

Returns: decoded topic names
Return type: [str]

tesp_support.fncs.get_keys_size(): Get the size of the keys

tesp_support.fncs.get_name()

Find the FNCS simulator name

Returns: the name of this simulator as provided in the ZPL or YAML file
Return type: str

tesp_support.fncs.get_simulator_count(): Find the FNCS simulator count

tesp_support.fncs.get_value(key)

Extract value from a FNCS message

Parameters: key (str) – the topic
Returns: decoded value
Return type: str

tesp_support.fncs.get_value_at(key, i)

For list publications, get the value by index

Parameters

key (str) – the topic
i (int) – the list index number

Returns

decoded value

Return type

str

tesp_support.fncs.get_values(key)

For list publications, get the list of values

Parameters: key (str) – the topic
Returns: decoded values
Return type: [str]

tesp_support.fncs.get_values_size(key)

For list publications, find how many values were published

Parameters: key (str) – the topic
Returns: the number of values for this topic
Return type: int

tesp_support.fncs.get_version()

Find the FNCS version

Returns: major, minor and patch numbers
Return type: int, int, int

tesp_support.fncs.initialize(config=None)

Initialize the FNCS configuration

Parameters: config (str) – a ZPL file. If None (default), provide YAML file in FNCS_CONFIG_FILE environment variable.

tesp_support.fncs.is_initialized()

Determine whether the FNCS library has been initialized

Returns: True if initialized, False if not.
Return type: Boolean

tesp_support.fncs.publish(key, value)

Publish a value over FNCS, under the simulator name

Parameters

key (str) – topic under the simulator name
value (str) – value

tesp_support.fncs.publish_anon(key, value)

Publish a value over FNCS, under the ‘anonymous’ simulator name

Parameters

key (str) – topic under ‘anonymous’
value (str) – value

tesp_support.fncs.route(sender, receiver, key, value)

Route a value over FNCS from sender to receiver

Parameters

sender (str) – simulator routing the message
receiver (str) – simulator to route the message to
key (str) – topic under the simulator name
value (str) – value

tesp_support.fncs.time_request(time)

FNCS time request

Parameters: time (int) – requested time.

tesp_support.fncs.update_time_delta(delta)

Update simulator time delta value

Parameters: delta (int) – time delta.

tesp_support.tso_PYPOWER module

tesp_support.glm_dict module

Functions to create metadata from a GridLAB-D input (GLM) file

Metadata is written to a JSON file, for convenient loading into a Python dictionary. It can be used for agent configuration, e.g., to initialize a forecasting model based on some nominal data. It’s also used with metrics output in post-processing.

Public Functions:

glm_dict: Writes the JSON metadata file.

tesp_support.glm_dict.append_include_file(lines, fname)

tesp_support.glm_dict.ercotMeterName(objname)

Enforces the meter naming convention for ERCOT

Replaces anything after the last _ with mtr.

Parameters: objname (str) – the GridLAB-D name of a house or inverter
Returns: The GridLAB-D name of upstream meter
Return type: str

tesp_support.glm_dict.glm_dict(nameroot, ercot=False, te30=False)

Writes the JSON metadata file from a GLM file

This function reads nameroot.glm and writes nameroot_glm_dict.json The GLM file should have some meters and triplex_meters with the bill_mode attribute defined, which identifies them as billing meters that parent houses and inverters. If this is not the case, ERCOT naming rules can be applied to identify billing meters.

Parameters

nameroot (str) – path and file name of the GLM file, without the extension
ercot (boolean) – request ERCOT billing meter naming. Defaults to false.
te30 (boolean) – request hierarchical meter handling in the 30-house test harness. Defaults to false.

tesp_support.glm_dict.isCommercialHouse(house_class)

tesp_support.glm_dict.ti_enumeration_string(tok): if thermal_integrity_level is an integer, convert to a string for the metadata

tesp_support.helpers module

Utility functions for use within tesp_support, including new agents.

class tesp_support.helpers.ClearingType(value)

Bases: IntEnum

Describes the market clearing type

BUYER = 5

EXACT = 3

FAILURE = 1

NULL = 0

PRICE = 2

SELLER = 4

class tesp_support.helpers.HelicsMsg(name, period)

Bases: object

config(_n, _v)

pubs(_g, _k, _t, _o, _p)

pubs_e(_g, _k, _t, _u)

pubs_n(_g, _k, _t)

subs(_k, _t, _o, _p)

subs_e(_r, _k, _t, _i)

subs_n(_k, _t)

write_file(_fn)

tesp_support.helpers.aggregate_bid(crv)

aggregates the buyer curve into a quadratic or straight-line fit with zero intercept

Parameters: crv (curve) – the accumulated buyer bids
Returns: Qunresp, Qmaxresp, degree, c2 and c1 scaled to MW instead of kW. c0 is always zero.
Return type: [float, float, int, float, float]

class tesp_support.helpers.curve

Bases: object

Accumulates a set of price, quantity bids for later aggregation. The default order is descending by price.

price

array of prices, in $/kWh

Type: [float]

quantity

array of quantities, in kW

Type: [float]

count

the number of collected bids

Type: int

total

the total kW bidding

Type: float

total_on

the total kW bidding that are currently on

Type: float

total_off

the total kW bidding that are currently off

Type: float

add_to_curve(price, quantity, is_on)

Add one point to the curve

Parameters

price (float) – the bid price, should be $/kWhr
quantity (float) – the bid quantity, should be kW
is_on (Boolean) – True if the load is currently on, False if not

set_curve_order(flag)

Set the curve order (by price) to ascending or descending

Parameters: flag (str) – ‘ascending’ or ‘descending’

tesp_support.helpers.gld_strict_name(val)

Sanitizes a name for GridLAB-D publication to FNCS GridLAB-D name should not begin with a number, or contain ‘-’ for FNCS

Parameters: val (str) – the input name
Returns: val with all ‘-’ replaced by ‘_’, and any leading digit replaced by ‘gld_’
Return type: str

tesp_support.helpers.idf_int(val)

Helper function to format integers for the EnergyPlus IDF input data file

Parameters: val (int) – the integer to format
Returns: the integer in string format, padded with a comma and zero or one blanks, in order to fill three spaces
Return type: str

tesp_support.helpers.parse_helic_input(arg)

Helper function to find the magnitude of a possibly complex number from Helics as a string

Parameters: arg (str) – The Helics value

tesp_support.helpers.parse_kva(arg)

Parse the kVA magnitude from GridLAB-D P+jQ volt-amperes in rectangular form

Parameters: arg (str) – the GridLAB-D P+jQ value
Returns: the parsed kva value
Return type: float

tesp_support.helpers.parse_kva_old(arg)

tesp_support.helpers.parse_kw(arg)

Parse the kilowatt load of a possibly complex number from FNCS

Parameters: arg (str) – the FNCS string value
Returns: the parsed number in kW, or 0 if parsing fails
Return type: float

tesp_support.helpers.parse_magnitude(arg)

Parse the magnitude of a possibly complex number from FNCS

Parameters: arg (str) – the FNCS string value
Returns: the parsed number, or 0 if parsing fails
Return type: float

tesp_support.helpers.parse_magnitude_1(arg)

Parse the magnitude of a possibly complex number from FNCS :param arg: the FNCS string value :type arg: str

Returns: the parsed number, or 0 if parsing fails
Return type: float

tesp_support.helpers.parse_magnitude_2(arg)

Helper function to find the magnitude of a possibly complex number from FNCS

Parameters: arg (str) – The FNCS value

tesp_support.helpers.parse_mva(arg)

Helper function to parse P+jQ from a FNCS value

Parameters: arg (str) – FNCS value in rectangular format
Returns: P [MW] and Q [MVAR]
Return type: float, float

tesp_support.helpers.parse_number(arg)

Parse floating-point number from a FNCS message; must not have leading sign or exponential notation

Parameters: arg (str) – the FNCS string value
Returns: the parsed number
Return type: float

tesp_support.helpers.zoneMeterName(ldname)

Enforces the meter naming convention for commercial zones The commercial zones must be children of load objects This routine replaces “_load_” with “_meter”.

Parameters: ldname (str) – the GridLAB-D name of a load, ends with _load_##
Returns: The GridLAB-D name of upstream meter
Return type: str

tesp_support.hvac module

Class that controls the responsive thermostat for one house.

Implements the ramp bidding method, with HVAC power as the bid quantity, and thermostat setting changes as the response mechanism.

class tesp_support.hvac.hvac(dict, key, aucObj)

Bases: object

This agent manages thermostat setpoint and bidding for a house

Parameters

dict (dict) – dictionary row for this agent from the JSON configuration file
key (str) – name of this agent, also key for its dictionary row
aucObj (simple_auction) – the auction this agent bids into

name

name of this agent

Type: str

control_mode

control mode from dict (CN_RAMP or CN_NONE, which still implements the setpoint schedule)

Type: str

houseName

name of the corresponding house in GridLAB-D, from dict

Type: str

meterName

name of the corresponding triplex_meter in GridLAB-D, from dict

Type: str

period

market clearing period, in seconds, from dict

Type: float

wakeup_start

hour of the day (0..24) for scheduled weekday wakeup period thermostat setpoint, from dict

Type: float

daylight_start

hour of the day (0..24) for scheduled weekday daytime period thermostat setpoint, from dict

Type: float

evening_start

hour of the day (0..24) for scheduled weekday evening (return home) period thermostat setpoint, from dict

Type: float

night_start

hour of the day (0..24) for scheduled weekday nighttime period thermostat setpoint, from dict

Type: float

wakeup_set

preferred thermostat setpoint for the weekday wakeup period, in deg F, from dict

Type: float

daylight_set

preferred thermostat setpoint for the weekday daytime period, in deg F, from dict

Type: float

evening_set

preferred thermostat setpoint for the weekday evening (return home) period, in deg F, from dict

Type: float

night_set

preferred thermostat setpoint for the weekday nighttime period, in deg F, from dict

Type: float

weekend_day_start

hour of the day (0..24) for scheduled weekend daytime period thermostat setpoint, from dict

Type: float

weekend_day_set

preferred thermostat setpoint for the weekend daytime period, in deg F, from dict

Type: float

weekend_night_start

hour of the day (0..24) for scheduled weekend nighttime period thermostat setpoint, from dict

Type: float

weekend_night_set

preferred thermostat setpoint for the weekend nighttime period, in deg F, from dict

Type: float

deadband

thermostat deadband in deg F, invariant, from dict

Type: float

offset_limit

maximum allowed change from the time-scheduled setpoint, in deg F, from dict

Type: float

ramp

bidding ramp denominator in multiples of the price standard deviation, from dict

Type: float

price_cap

the highest allowed bid price in $/kwh, from dict

Type: float

bid_delay

from dict, not implemented

Type: float

use_predictive_bidding

from dict, not implemented

Type: float

std_dev

standard deviation of expected price, determines the bidding ramp slope, initialized from aucObj

Type: float

mean

mean of the expected price, determines the bidding ramp origin, initialized from aucObj

Type: float

Trange

the allowed range of setpoint variation, bracketing the preferred time-scheduled setpoint

Type: float

air_temp

current air temperature of the house in deg F

Type: float

hvac_kw

most recent non-zero HVAC power in kW, this will be the bid quantity

Type: float

mtr_v

current line-neutral voltage at the triplex meter

Type: float

hvac_on

True if the house HVAC is currently running

Type: Boolean

basepoint

the preferred time-scheduled thermostat setpoint in deg F

Type: float

setpoint

the thermostat setpoint, including price response, in deg F

Type: float

bid_price

the current bid price in $/kwh

Type: float

cleared_price

the cleared market price in $/kwh

Type: float

bid_accepted()

Update the thermostat setting if the last bid was accepted

The last bid is always “accepted”. If it wasn’t high enough, then the thermostat could be turned up.p

Returns: True if the thermostat setting changes, False if not.
Return type: Boolean

change_basepoint(hod, dow)

Updates the time-scheduled thermostat setting

Parameters

hod (float) – the hour of the day, from 0 to 24
dow (int) – the day of the week, zero being Monday

Returns

True if the setting changed, Falso if not

Return type

Boolean

formulate_bid()

Bid to run the air conditioner through the next period

Returns: bid price in $/kwh, bid quantity in kW and current HVAC on state, or None if not bidding
Return type: [float, float, Boolean]

inform_bid(price)

Set the cleared_price attribute

Parameters: price (float) – cleared price in $/kwh

set_air_temp_from_fncs_str(val)

Sets the air_temp attribute

Parameters: val (str) – FNCS message with temperature in degrees Fahrenheit

set_air_temp_from_helics(val)

set_hvac_load_from_fncs_str(val)

Sets the hvac_load attribute, if greater than zero

Parameters: val (str) – FNCS message with load in kW

set_hvac_load_from_helics(val)

set_hvac_state_from_fncs_str(val)

Sets the hvac_on attribute

Parameters: val (str) – FNCS message with state, ON or OFF

set_hvac_state_from_helics(val)

set_voltage_from_fncs_str(val)

Sets the mtr_v attribute

Parameters: val (str) – FNCS message with meter line-neutral voltage

set_voltage_from_helics(val)

tesp_support.make_ems module

Creates and merges the EMS for an EnergyPlus building model

Public Functions:

make_ems: Creates the energy management system (EMS) for FNCS/HELICS to interface with EnergyPlus.
merge_idf: Assembles the base IDF, the EMS, start time and end time

tesp_support.make_ems.cooling_coil_sensor(name, target, op)

tesp_support.make_ems.get_eplus_token(sval)

tesp_support.make_ems.global_variable(name, op)

tesp_support.make_ems.heating_coil_sensor(name, target, op)

tesp_support.make_ems.make_ems(sourcedir='./output', baseidf='SchoolBase.idf', target='ems.idf', write_summary=False, bHELICS=False)

Creates the EMS for an EnergyPlus building model

Parameters

target (str) – desired output file in PWD, default ems.idf
baseidf (str) – is the original EnergyPlus model file without the EMS
sourcedir (str) – directory of the output from EnergyPlus baseline simulation, default ./output

tesp_support.make_ems.merge_idf(base, ems, StartTime, EndTime, target, StepsPerHour)

Assembles a base EnergyPlus building model with EMS and simulation period

Parameters

base (str) – fully qualified base IDF model
ems (str) – fully qualified EMS model file
StartTime (str) – Date-Time to start simulation, Y-m-d H:M:S
EndTime (str) – Date-Time to end simulation, Y-m-d H:M:S

tesp_support.make_ems.output_variable(name, target, op)

tesp_support.make_ems.print_idf_summary(target, zones, zonecontrols, thermostats, schedules, hcoils, ccoils, hvacs)

tesp_support.make_ems.schedule_actuator(name, target, op)

tesp_support.make_ems.schedule_sensor(name, op)

tesp_support.make_ems.summarize_idf(fname, baseidf)

tesp_support.make_ems.valid_var(name)

tesp_support.make_ems.write_new_ems(target, zones, zonecontrols, thermostats, schedules, hcoils, ccoils, hvacs, bHELICS)

tesp_support.make_ems.zone_heating_sensor(name, op)

tesp_support.make_ems.zone_occupant_sensor(name, op)

tesp_support.make_ems.zone_sensible_cooling_sensor(name, op)

tesp_support.make_ems.zone_sensible_heating_sensor(name, op)

tesp_support.make_ems.zone_temperature_sensor(name, op)

tesp_support.parse_msout module

tesp_support.parse_msout.next_matrix(fp, var)

tesp_support.parse_msout.next_val(fp, var, bInteger=True)

tesp_support.parse_msout.read_most_solution(fname='msout.txt')

tesp_support.precool module

Classes for NIST TE Challenge 2 example

The precool_loop class manages time stepping and FNCS messages for the precooler agents, which adjust thermostat setpoints in response to time-of-use rates and overvoltages. The precooler agents also estimate house equivalent thermal parameter (ETP) models based on total floor area, number of stories, number of exterior doors and and estiamted thermal integrity level. This ETP estimate serves as an example for other agent developers; it’s not actually used by the precooler agent.

Public Functions:

precooler_loop: Initializes and runs the precooler agents.

tesp_support.precool.precool_loop(nhours, metrics_root, dict_root, response='PriceVoltage')

Function that supervises FNCS messages and time stepping for precooler agents

Opens metrics_root_agent_dict.json and metrics_root_glm_dict.json for configuration. Writes precool_metrics_root.json at completion.

Parameters

nhours (float) – number of hours to simulate
metrics_root (str) – name of the case, without file extension
dict_root (str) – repeat metrics_root, or the name of a shared case dictionary without file extension
response (str) – combination of Price and/or Voltage

class tesp_support.precool.precooler(name, agentrow, gldrow, k, mean, stddev, lockout_time, precooling_quiet, precooling_off, bPrice, bVoltage)

Bases: object

This agent manages the house thermostat for time-of-use and overvoltage responses.

References

NIST TE Modeling and Simulation Challenge

Parameters

name (str) – name of this agent
agentrow (dict) – row from the FNCS configuration dictionary for this agent
gldrow (dict) – row from the GridLAB-D metadata dictionary for this agent’s house
k (float) – bidding function denominator, in multiples of stddev
mean (float) – mean of the price
stddev (float) – standard deviation of the price
lockout_time (float) – time in seconds between allowed changes due to voltage
precooling_quiet (float) – time of day in seconds when precooling is allowed
precooling_off (float) – time of day in seconds when overvoltage precooling is always turned off

name

name of this agent

Type: str

meterName

name of the corresponding triplex_meter in GridLAB-D, from agentrow

Type: str

night_set

preferred thermostat setpoint during nighttime hours, deg F, from agentrow

Type: float

day_set

preferred thermostat setpoint during daytime hours, deg F, from agentrow

Type: float

day_start_hour

hour of the day when daytime thermostat setting period begins, from agentrow

Type: float

day_end_hour

hour of the day when daytime thermostat setting period ends, from agentrow

Type: float

deadband

thermostat deadband in deg F, invariant, from agentrow, from agentrow

Type: float

vthresh

meter line-to-neutral voltage that triggers precooling, from agentrow

Type: float

toffset

temperature setpoint change for precooling, in deg F, from agentrow

Type: float

k

bidding function denominator, in multiples of stddev

Type: float

mean

mean of the price

Type: float

stddev

standard deviation of the price

Type: float

lockout_time

time in seconds between allowed changes due to voltage

Type: float

precooling_quiet

time of day in seconds when precooling is allowed

Type: float

precooling_off

time of day in seconds when overvoltage precooling is always turned off

Type: float

air_temp

current air temperature of the house in deg F

Type: float

mtr_v

current line-neutral voltage at the triplex meter

Type: float

basepoint

the preferred time-scheduled thermostat setpoint in deg F

Type: float

setpoint

the thermostat setpoint, including price response, in deg F

Type: float

lastchange

time of day in seconds when the setpoint was last changed

Type: float

precooling

True if the house is precooling, False if not

Type: Boolean

ti

thermal integrity level, as enumerated for GridLAB-D, from gldrow

Type: int

sqft (float: total floor area in square feet, from gldrow

stories

number of stories, from gldrow

Type: int

doors

number of exterior doors, from gldrow

Type: int

UA

heat loss coefficient

Type: float

CA

total air thermal mass

Type: float

HM

interior mass surface conductance

Type: float

CM

total house thermal mass

Type: float

check_setpoint_change(hour_of_day, price, time_seconds)

Update the setpoint for time of day and price

Parameters

hour_of_day (float) – the current time of day, 0..24
price (float) – the current price in $/kwh
time_seconds (long long) – the current FNCS time in seconds

Returns

True if the setpoint changed, False if not

Return type

Boolean

get_temperature_deviation()

For metrics, find the difference between air temperature and time-scheduled (preferred) setpoint

Returns: absolute value of deviation
Return type: float

make_etp_model()

Sets the ETP parameters from configuration data

References

Thermal Integrity Table Inputs and Defaults

set_air_temp(val)

Set the air_temp member variable

Parameters: val (str) – FNCS message with temperature in degrees Fahrenheit

set_voltage(val)

Sets the mtr_v attribute

Parameters: val (str) – FNCS message with meter line-neutral voltage

tesp_support.prep_eplus module

tesp_support.prep_eplus.configure_eplus(caseConfig, template_dir)

tesp_support.prep_eplus.make_gld_eplus_case(fname, bGlmReady=False)

tesp_support.prep_eplus.prepare_bldg_dict(caseConfig)

tesp_support.prep_eplus.prepare_glm_dict(caseConfig)

tesp_support.prep_eplus.prepare_glm_file(caseConfig)

tesp_support.prep_eplus.prepare_glm_helics(caseConfig, fedMeters, fedLoadNames)

tesp_support.prep_eplus.prepare_run_script(caseConfig, fedMeters)

tesp_support.prep_eplus.writeGlmClass(theseLines, thisClass, op)

tesp_support.prep_precool module

Writes the precooling agent and GridLAB-D metadata for NIST TE Challenge 2 example

Public Functions:

prep_precool: writes the JSON and YAML files

tesp_support.prep_precool.prep_precool(nameroot, time_step=15)

Sets up agent configurations for the NIST TE Challenge 2 example

Reads the GridLAB-D data from nameroot.glm; it should contain houses with thermal_integrity_level attributes. Writes:

nameroot_agent_dict.json, contains configuration data for the precooler agents
nameroot_precool.yaml, contains FNCS subscriptions for the precooler agents
nameroot_FNCS_Config.txt, a GridLAB-D include file with FNCS publications and subscriptions

Parameters

nameroot (str) – the name of the GridLAB-D file, without extension
time_step (int) – time step period

tesp_support.prep_substation module

Sets up the FNCS and agent configurations for te30 and sgip1 examples

This works for other TESP cases that have one GridLAB-D file, one EnergyPlus model, and one PYPOWER model. Use tesp_case or tesp_config modules to specify supplemental configuration data for these TESP cases, to be provided as the optional jsonfile argument to prep_substation.

Public Functions:

prep_substation: processes a GridLAB-D file for one substation and one or more feeders

tesp_support.prep_substation.ProcessGLM(fileroot)

Helper function that processes one GridLAB-D file

Reads fileroot.glm and writes:

fileroot_agent_dict.json, contains configuration data for the simple_auction and hvac agents
fileroot_substation.yaml, contains FNCS subscriptions for the psimple_auction and hvac agents
nameroot_FNCS_Config.txt, a GridLAB-D include file with FNCS publications and subscriptions
fileroot_HELICS_substation.json, contains HELICS subscriptions for the psimple_auction and hvac agents
nameroot_HELICS_gld_msg.json, a GridLAB-D include file with HELICS publications and subscriptions

Parameters: fileroot (str) – path to and base file name for the GridLAB-D file, without an extension

tesp_support.prep_substation.prep_substation(gldfileroot, jsonfile='', bus_id=None)

Process a base GridLAB-D file with supplemental JSON configuration data

Always reads gldfileroot.glm and writes:

gldfileroot_agent_dict.json, contains configuration data for the simple_auction and hvac agents
gldfileroot_substation.yaml, contains FNCS subscriptions for the psimple_auction and hvac agents
gldfileroot_FNCS_Config.txt, a GridLAB-D include file with FNCS publications and subscriptions
gldfileroot_Weather_Config.json, contains configuration data for the weather agent

If provided, this function also reads jsonfile as created by tesp_config and used by tesp_case. This supplemental data includes time-scheduled thermostat setpoints (NB: do not use the scheduled setpoint feature within GridLAB-D, as the first FNCS messages will erase those schedules during simulation). The supplemental data also includes time step and market period, the load scaling factor to PYPOWER, ramp bidding function parameters and the EnergyPlus connection point. If not provided, the default values from te30 and sgip1 examples will be used.

Parameters

bus_id – substation bus identifier
gldfileroot (str) – path to and base file name for the GridLAB-D file, without an extension
jsonfile (str) – fully qualified path to an optional JSON configuration file (if not provided, an E+ connection to Eplus_load will be created)

tesp_support.process_agents module

Functions to plot data from GridLAB-D substation agents

Public Functions:

process_agents: Reads the data and metadata, then makes the plots.

tesp_support.process_agents.plot_agents(dict, save_file=None, save_only=False)

tesp_support.process_agents.process_agents(nameroot, dictname='', save_file=None, save_only=False, print_dictionary=False)

Plots cleared price, plus bids from the first HVAC controller

This function reads auction_nameroot_metrics.json and controller_nameroot_metrics.json for the data; it reads nameroot_glm_dict.json for the metadata. These must all exist in the current working directory. Makes one graph with 2 subplots:

Cleared price from the only auction, and bid price from the first controller
Bid quantity from the first controller

Parameters

nameroot (str) – name of the TESP case, not necessarily the same as the GLM case, without the extension
dictname (str) – metafile name (with json extension) for a different GLM dictionary, if it’s not nameroot_glm_dict.json. Defaults to empty.
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_agents.read_agent_metrics(path, nameroot, dictname='', print_dictionary=False)

tesp_support.process_eplus module

Functions to plot data from the EnergyPlus agent

Public Functions:

process_eplus: Reads the data and metadata, then makes the plots.

tesp_support.process_eplus.plot_eplus(dict, title=None, save_file=None, save_only=False)

tesp_support.process_eplus.process_eplus(nameroot, title=None, save_file=None, save_only=False)

Plots the min and max line-neutral voltages for every billing meter

This function reads eplus_nameroot_metrics.json for both metadata and data. This must exist in the current working directory. One graph is generated with 3 subplots:

Cooling system setpoint, actual temperature and the difference between them.
Heating system setpoint, actual temperature and the difference between them.
Price that the building controller responded to.

Parameters

nameroot (str) – name of the TESP case, not necessarily the same as the EnergyPlus case, without the extension
title (str) – supertitle for the page of plots.
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_eplus.read_eplus_metrics(path, nameroot, quiet=False)

tesp_support.process_gld module

Functions to plot data from GridLAB-D

Public Functions:

process_gld: Reads the data and metadata, then makes the plots.

tesp_support.process_gld.plot_gld(dict, save_file=None, save_only=False)

tesp_support.process_gld.process_gld(nameroot, dictname='', save_file=None, save_only=False)

Plots a summary/sample of power, air temperature and voltage

This function reads substation_nameroot_metrics.json, billing_meter_nameroot_metrics.json and house_nameroot_metrics.json for the data; it reads nameroot_glm_dict.json for the metadata. These must all exist in the current working directory. Makes one graph with 4 subplots:

Substation real power and losses
Average air temperature over all houses
Min/Max line-to-neutral voltage and Min/Max line-to-line voltage at the first billing meter
Min, Max and Average air temperature at the first house

Parameters

nameroot (str) – name of the TESP case, not necessarily the same as the GLM case, without the extension
dictname (str) – metafile name (with json extension) for a different GLM dictionary, if it’s not nameroot_glm_dict.json. Defaults to empty.
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_gld.read_gld_metrics(path, nameroot, dictname='')

tesp_support.process_houses module

Functions to plot house data from GridLAB-D

Public Functions:

process_houses: Reads the data and metadata, then makes the plot.

tesp_support.process_houses.plot_houses(dict, save_file=None, save_only=False)

tesp_support.process_houses.process_houses(nameroot, dictname='', save_file=None, save_only=True)

Plots the temperature and HVAC power for every house

This function reads substation_nameroot_metrics.json and house_nameroot_metrics.json for the data; it reads nameroot_glm_dict.json for the metadata. These must all exist in the current working directory. Makes one graph with 2 subplots:

Average air temperature at every house
Average HVAC power at every house

Parameters

nameroot (str) – name of the TESP case, not necessarily the same as the GLM case, without the extension
dictname (str) – metafile name (with json extension) for a different GLM dictionary, if it’s not nameroot_glm_dict.json. Defaults to empty.
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_houses.read_house_metrics(path, nameroot, dictname='')

tesp_support.process_inv module

Functions to plot inverter and volt-var data from GridLAB-D, for NIST TE Challenge 2

Public Functions:

process_inv: Reads the data and metadata, then makes the plots.

tesp_support.process_inv.plot_inv(plt, save_file=None, save_only=False)

tesp_support.process_inv.process_inv(nameroot, dictname='', save_file=None, save_only=False)

Plots inverter and volt-var data for the NIST TE Challenge 2 / IEEE 8500 examples

This function reads substation_nameroot_metrics.json, billing_meter_nameroot_metrics.json, capacitor_nameroot_metrics.json, regulator_nameroot_metrics.json, house_nameroot_metrics.json and inverter_nameroot_metrics.json for the data; it reads nameroot_glm_dict.json for the metadata. If possible, it reads precool_nameroot_metrics.json for temperature deviation. These must all exist in the current working directory. One graph is generated with 10 subplots:

Average P and Q over all inverters
Min, Max and Average line-neutral voltage over all billing meters
Average air temperature over all houses
Average temperature deviations from the setpoint over all houses
Total of ANSI C84 A and B range violation counts, summing over all billing meters
Total of ANSI C84 A and B range violation durations, summing over all billing meters
Substation total power, losses, house power, house HVAC power and house waterheater power
The accumulated bill, summed over all billing meters
The accumulated capacitor switching counts for each of 4 capacitor banks, if found, as in the IEEE 8500 case
The accumulated regulator counts for each of 4 voltage regulators, if found, as in the IEEE 8500 case

Parameters

nameroot (str) – name of the TESP case, not necessarily the same as the GLM case, without the extension
dictname (str) – metafile name (with json extension) for a different GLM dictionary, if it’s not nameroot_glm_dict.json. Defaults to empty.
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_inv.process_ninv(nameroot, title=None, save_file=None, save_only=False)

Plots

Parameters

nameroot (str) – file name of the TESP case, not necessarily the same as the PYPOWER case, without the JSON extension
title (str) –
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_inv.read_inv_metrics(path, nameroot, dictname='', save_file=None, save_only=False)

Plots inverter and volt-var data for the NIST TE Challenge 2 / IEEE 8500 examples

This function reads substation_nameroot_metrics.json, billing_meter_nameroot_metrics.json, capacitor_nameroot_metrics.json, regulator_nameroot_metrics.json, house_nameroot_metrics.json and inverter_nameroot_metrics.json for the data; it reads nameroot_glm_dict.json for the metadata. If possible, it reads precool_nameroot_metrics.json for temperature deviation. These must all exist in the current working directory. One graph is generated with 10 subplots:

Average P and Q over all inverters
Min, Max and Average line-neutral voltage over all billing meters
Average air temperature over all houses
Average temperature deviations from the setpoint over all houses
Total of ANSI C84 A and B range violation counts, summing over all billing meters
Total of ANSI C84 A and B range violation durations, summing over all billing meters
Substation total power, losses, house power, house HVAC power and house waterheater power
The accumulated bill, summed over all billing meters
The accumulated capacitor switching counts for each of 4 capacitor banks, if found, as in the IEEE 8500 case
The accumulated regulator counts for each of 4 voltage regulators, if found, as in the IEEE 8500 case

Parameters

nameroot (str) – name of the TESP case, not necessarily the same as the GLM case, without the extension
dictname (str) – metafile name (with json extension) for a different GLM dictionary, if it’s not nameroot_glm_dict.json. Defaults to empty.
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_pypower module

Functions to plot bus and generator data from PYPOWER

Public Functions:

process_pypower: Reads the data and metadata, then makes the plots.

tesp_support.process_pypower.plot_pypower(dict, title=None, save_file=None, save_only=False)

tesp_support.process_pypower.process_pypower(nameroot, title=None, save_file=None, save_only=True)

Plots bus and generator quantities for the 9-bus system used in te30 or sgip1 examples

This function reads bus_nameroot_metrics.json and gen_nameroot_metrics.json for the data, and nameroot_m_dict.json for the metadata. These must all exist in the current working directory. One graph is generated with 8 subplots:

Bus P and Q demands, at the single GridLAB-D connection
Bus P and Q locational marginal prices (LMP), at the single GridLAB-D connection
Bus Vmin, Vmax and Vavg, at the single GridLAB-D connection
All 4 generator prices
Generator 1 P and Q output
Generator 2 P and Q output
Generator 3 P and Q output
Generator 4 P and Q output

Parameters

nameroot (str) – file name of the TESP case, not necessarily the same as the PYPOWER case, without the JSON extension
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_pypower.read_pypower_metrics(path, nameroot)

tesp_support.process_voltages module

Functions to plot all billing meter voltages from GridLAB-D

Public Functions:

process_voltages: Reads the data and metadata, then makes the plot.

tesp_support.process_voltages.plot_voltages(dict, save_file=None, save_only=False)

tesp_support.process_voltages.process_voltages(nameroot, dictname='', save_file=None, save_only=True)

Plots the min and max line-neutral voltages for every billing meter

This function reads substation_nameroot_metrics.json and billing_meter_nameroot_metrics.json for the voltage data, and nameroot_glm_dict.json for the meter names. These must all exist in the current working directory. One graph is generated with 2 subplots:

The Min line-to-neutral voltage at each billing meter
The Max line-to-neutral voltage at each billing meter

Parameters

nameroot (str) – name of the TESP case, not necessarily the same as the GLM case, without the extension
dictname (str) – metafile name (with json extension) for a different GLM dictionary, if it’s not nameroot_glm_dict.json. Defaults to empty.
save_file (str) – name of a file to save plot, should include the png or pdf extension to determine type.
save_only (Boolean) – set True with save_file to skip the display of the plot. Otherwise, script waits for user keypress.

tesp_support.process_voltages.read_voltage_metrics(path, nameroot, dictname='')

tesp_support.run_tesp_case module

Auto test runner for TESP run* cases Runs a test case based on pre-existing shell script file.

If FNCS or HELICS broker exist the test waits for: the broker process to finish before function returns.

This code has limited functionality as the ‘run*’ scripts for the examples are written in a very specified way.

tesp_support.run_tesp_case.block_test(call)

tesp_support.run_tesp_case.init_tests()

tesp_support.run_tesp_case.process_line(line, local_vars)

tesp_support.run_tesp_case.report_tests()

tesp_support.run_tesp_case.run_test(file_name, casename=None)

tesp_support.run_tesp_case.start_test(casename=None)

tesp_support.simple_auction module

Double-auction mechanism for the 5-minute markets in te30 and sgip1 examples

The substation_loop module manages one instance of this class per GridLAB-D substation.

Todo

Initialize and update price history statistics
Allow for adjustment of clearing_scalar
Handle negative price bids from HVAC agents, currently they are discarded
Distribute marginal quantities and fractions; these are not currently applied to HVACs

2021-10-29 TDH: Key assumptions we need to refactor out of this: - This is a retail market working under a wholesale market. We want something more general that can operate at any market level. - Load state is not necessary information to run a market. PNNL TSP has traditionally had the construct of responsive and unresponsive loads and I think the idea is crucial for being able to get good quadratic curve fits on the demand curve (by lopping off the unresponsive portion) but I think we should refactor this so that these assumptions are not so tightly integrated with the formulation.

class tesp_support.simple_auction.simple_auction(dict, key)

Bases: object

This class implements a simplified version of the double-auction market embedded in GridLAB-D.

References

Market Module Overview - Auction

Parameters

dict (dict) – a row from the agent configuration JSON file
key (str) – the name of this agent, which is the market key from the agent configuration JSON file

name

the name of this auction, also the market key from the configuration JSON file

Type: str

std_dev

the historical standard deviation of the price, in $/kwh, from dict

Type: float

mean

the historical mean price in $/kwh, from dict

Type: float

pricecap

the maximum allowed market clearing price, in $/kwh, from dict

Type: float

max_capacity_reference_bid_quantity

this market’s maximum capacity, likely defined by a physical limitation in the circuit(s) being managed.

Type: float

statistic_mode

always 1, not used, from dict

Type: int

stat_mode

always ST_CURR, not used, from dict

Type: str

stat_interval

always 86400 seconds, for one day, not used, from dict

Type: str

stat_type

always mean and standard deviation, not used, from dict

Type: str

stat_value

always zero, not used, from dict

Type: str

curve_buyer

data structure to accumulate buyer bids

Type: curve

curve_seller

data structure to accumulate seller bids

Type: curve

refload

the latest substation load from GridLAB-D. This is initially assumed to be all unresponsive and using the load state parameter when adding demand bids (which are generally price_responsive) all loads that bid and are on are removed from the assumed unresponsive load value

Type: float

lmp

the latest locational marginal price from the bulk system market

Type: float

unresp

unresponsive load, i.e., total substation load less the bidding, running HVACs

Type: float

agg_unresp

aggregated unresponsive load, i.e., total substation load less the bidding, running HVACs

Type: float

agg_resp_max

total load of the bidding HVACs

Type: float

agg_deg

degree of the aggregate bid curve polynomial, should be 0 (zero or one bids), 1 (2 bids) or 2 (more bids)

Type: int

agg_c2

second-order coefficient of the aggregate bid curve

Type: float

agg_c1

first-order coefficient of the aggregate bid curve

Type: float

clearing_type

describes the solution type or boundary case for the latest market clearing

Type: helpers.ClearingType

clearing_quantity

quantity at the last market clearing

Type: float

clearing_price

price at the last market clearing

Type: float

marginal_quantity

quantity of a partially accepted bid

Type: float

marginal_frac

fraction of the bid quantity accepted from a marginal buyer or seller

Type: float

clearing_scalar

used for interpolation at boundary cases, always 0.5

Type: float

add_unresponsive_load(quantity)

aggregate_bids(): Aggregates the unresponsive load and responsive load bids for submission to the bulk system market

clear_bids(): Re-initializes curve_buyer and curve_seller, sets the unresponsive load estimate to the total substation load.

clear_market(tnext_clear=0, time_granted=0)

Solves for the market clearing price and quantity

Uses the current contents of curve_seller and curve_buyer. Updates clearing_price, clearing_quantity, clearing_type, marginal_quantity and marginal_frac.

Parameters

tnext_clear (int) – next clearing time in seconds, should be <= time_granted, for the log file only
time_granted (int) – the current time in seconds, for the log file only

collect_bid(bid)

Gather HVAC bids into curve_buyer

Also adjusts the unresponsive load estimate, by subtracting the HVAC power if the HVAC is on.

Parameters: bid ([float, float, Boolean]) – price in $/kwh, quantity in kW and the HVAC on state

initAuction()

Sets the clearing_price and lmp to the mean price

2021-10-29 TDH: TODO - Any reason we can’t put this in constructor?

set_lmp(lmp)

Sets the lmp attribute

Parameters: lmp (float) – locational marginal price from the bulk system market

set_refload(kw)

Sets the refload attribute

Parameters: kw (float) – GridLAB-D substation load in kw

supplier_bid(bid)

Gather supplier bids into curve_seller

Use this to enter curves in step-wise blocks.

Parameters: bid ([float, float]) – price in $/kwh, quantity in kW

surplusCalculation(tnext_clear=0, time_granted=0)

Calculates consumer surplus (and its average) and supplier surplus.

This function goes through all the bids higher than clearing price from buyers to calculate consumer surplus, and also accumlates the quantities that will be cleared while doing so. Of the cleared quantities, the quantity for unresponsive loads are also collected. Then go through each seller to calculate supplier surplus. Part of the supplier surplus corresponds to unresponsive load are excluded and calculated separately.

Parameters

tnext_clear (int) – next clearing time in seconds, should be <= time_granted, for the log file only
time_granted (int) – the current time in seconds, for the log file only

Returns

None

update_statistics(): Update price history statistics - not implemented

tesp_support.substation module

Manages the simple_auction and hvac agents for the te30 and sgip1 examples

Public Functions:

substation_loop: initializes and runs the agents

Todo

Getting an overflow error when killing process - investigate whether that happens if simulation runs to completion
Allow changes in the starting date and time; now it’s always midnight on July 1, 2013
Allow multiple markets per substation, e.g., 5-minute and day-ahead for the DSO+T study

tesp_support.substation.fncs_substation_loop(configfile, metrics_root, hour_stop, flag)

Helper function that initializes and runs the agents

Reads configfile. Writes auction_metrics_root_metrics.json and controller_metrics_root_metrics.json upon completion.

Parameters

configfile (str) – fully qualified path to the JSON agent configuration file
metrics_root (str) – base name of the case for metrics output
hour_stop (float) – number of hours to simulation
flag (str) – WithMarket or NoMarket to use the simple_auction, or not

tesp_support.substation.helics_substation_loop(configfile, metrics_root, hour_stop, flag, helicsConfig)

tesp_support.substation.substation_loop(configfile, metrics_root, hour_stop=48, flag='WithMarket', helicsConfig=None)

Wrapper for inner_substation_loop

When inner_substation_loop finishes, timing and memory metrics will be printed for non-Windows platforms.

tesp_support.tesp_case module

Creates and fills a subdirectory with files to run a TESP simulation

Use tesp_config to graphically edit the case configuration

Public Functions:

make_tesp_case: sets up for a single-shot TESP case
make_monte_carlo_cases: sets up for a Monte Carlo TESP case of up to 20 shots
first_tesp_feeder: customization of make_tesp_case that will accept more feeders
add_tesp_feeder: add another feeder to the case directory created by first_tesp_feeder

tesp_support.tesp_case.add_tesp_feeder(cfgfile)

Wrapper function to start a single TESP case configuration.

This function opens the JSON file, and calls write_tesp_case for just the GridLAB-D files. The subdirectory targetdir doesn’t have to match the case name in cfgfile, and it should be created first with make_tesp_case

Parameters: cfgfile (str) – JSON file containing the TESP case configuration

tesp_support.tesp_case.make_monte_carlo_cases(cfgfile='test.json')

Writes up to 20 TESP simulation case setups to a directory for Monte Carlo simulations

Latin hypercube sampling is recommended; sample values may be specified via tesp_config

Parameters: cfgfile (str) – JSON file containing the TESP case configuration

tesp_support.tesp_case.make_tesp_case(cfgfile='test.json')

Wrapper function for a single TESP case configuration.

This function opens the JSON file, and calls write_tesp_case

Parameters: cfgfile (str) – JSON file containing the TESP case configuration

tesp_support.tesp_case.modify_mc_config(config, mcvar, band, sample)

Helper function that modifies the Monte Carlo configuration for a specific sample, i.e., shot

For variables that have a band associated, the agent preparation code will apply additional randomization. This applies to thermostat ramps, offset limits, and period starting or ending times. For those variables, the Monte Carlo sample value is a mean, and the agent preparation code will apply a uniform distribution to obtain the actual value for each house.

tesp_support.tesp_case.write_tesp_case(config, cfgfile, freshdir=True)

Writes the TESP case from data structure to JSON file

This function assumes one GridLAB-D, one EnergyPlus, one PYPOWER and one substation_loop federate will participate in the TESP simulation. See the DSO+T study functions, which are customized to ERCOT 8-bus and 200-bus models, for examples of other configurations.

The TESP support directories, working directory and case name are all specified in config. This function will create one directory as follows:

workdir = config[‘SimulationConfig’][‘WorkingDirectory’]

casename = config[‘SimulationConfig’][‘CaseName’]

new directory created will be casedir = workdir/casename

This function will read or copy several files that are specified in the config. They should all exist. These include taxonomy feeders, GridLAB-D schedules, weather files, a base EnergyPlus model, a base PYPOWER model, and supporting scripts for the end user to invoke from the casedir. The user could add more base model files weather files or schedule files under the TESP support directory, where this tesp_case module will be able to find and use them.

This function will launch and wait for 6 subprocesses to assist in the case configuration. All must execute successfully:

TMY3toTMY2_ansi, which converts the user-selected TMY3 file to TMY2

tesp.convert_tmy2_to_epw, which converts the TMY2 file to EPW for EnergyPlus

tesp.TMY3toCSV, which converts the TMY3 file to CSV for the weather agent

tesp.populate_feeder, which populates the user-selected taxonomy feeder with houses and DER

tesp.glm_dict, which creates metadata for the populated feeder

tesp.prep_substation, which creates metadata and FNCS configurations for the substation agents

As the configuration process finishes, several files are written to casedir:

Casename.glm: the GridLAB-D model, copied and modified from the TESP support directory

Casename_FNCS_Config.txt: FNCS subscriptions and publications, included by Casename.glm

Casename_agent_dict.json: metadata for the simple_auction and hvac agents

Casename_glm_dict.json: metadata for Casename.glm

Casename_pp.json: the PYPOWER model, copied and modified from the TESP support directory

Casename_substation.yaml: FNCS subscriptions and time step for the substation, which manages the simple_auction and hvac controllers

NonGLDLoad.txt: non-responsive load data for the PYPOWER model buses, currently hard-wired for the 9-bus model. See the ERCOT case files for examples of expanded options.

SchoolDualController.idf: the EnergyPlus model, copied and modified from the TESP support directory

WA-Yakima_Air_Terminal.epw: the selected weather file for EnergyPlus, others can be selected

WA-Yakima_Air_Terminal.tmy3: the selected weather file for GridLAB-D, others can be selected

appliance_schedules.glm: time schedules for GridLAB-D

clean.sh: Linux/Mac OS X helper to clean up simulation outputs

commercial_schedules.glm: non-responsive non-responsive time schedules for GridLAB-D, invariant

eplus.yaml: FNCS subscriptions and time step for EnergyPlus

eplus_agent.yaml: FNCS subscriptions and time step for the EnergyPlus agent

kill5570.sh: Linux/Mac OS X helper to kill all federates listening on port 5570

launch_auction.py: helper script for the GUI solution monitor to launch the substation federate

launch_pp.py: helper script for the GUI solution monitor to launch the PYPOWER federate

monitor.py: helper to launch the GUI solution monitor (FNCS_CONFIG_FILE envar must be set for this process, see gui.sh under examples/te30)

plots.py: helper script that will plot a selection of case outputs

pypower.yaml: FNCS subscriptions and time step for PYPOWER

run.sh: Linux/Mac OS X helper to launch the TESP simulation

tesp_monitor.json: shell commands and other configuration data for the solution monitor GUI

tesp_monitor.yaml: FNCS subscriptions and time step for the solution monitor GUI

water_and_setpoint_schedule_v5.glm: non-responsive time schedules for GridLAB-D, invariant

weather.dat: CSV file of temperature, pressure, humidity, solar direct, solar diffuse and wind speed

Parameters

config (dict) – the complete case data structure
cfgfile (str) – the name of the JSON file that was read
freshdir (boolean) – flag to create the directory and base files anew

Todo

Write gui.sh, per the te30 examples

tesp_support.tesp_config module

Presents a GUI to configure and package TESP cases

Public Functions:

show_tesp_config: Initializes and runs the GUI

References

Graphical User Interfaces with Tk

class tesp_support.tesp_config.TespConfigGUI(master)

Bases: object

Manages a seven-page GUI for case configuration

The GUI opens and saves a JSON file in the format used by tesp.tesp_config

Todo

Possible data loss if the user updated the number of Monte Carlo cases, but didn’t click the Update button before saving the case configuration.

nb

the top-level GUI with tabbed pages

Type: Notebook

f1

the page for date/time setup, along with T&D files, weather files and file paths

Type: Frame

f2

the page for feeder generator setup

Type: Frame

f3

the page for PYPOWER setup

Type: Frame

f4

the page for EnergyPlus setup

Type: Frame

f5

the page for simple_auction and hvac agent setup

Type: Frame

f6

the page for time-scheduled thermostat settings

Type: Frame

f7

the page for Monte Carlo setup

Type: Frame

AttachFrame(tag, vars)

Creates a GUI page and loads it with data

Label, Combobox and Entry (i.e. edit) controls are automatically created for each row of data

Parameters

tag (str) – the name of the Frame, i.e., GUI page
vars (dict) – the section of case configuration data to be loaded onto this new GUI page

InitializeMonteCarlo(n)

Makes default variable choices and then initializes the Monte Carlo GUI page

Parameters: n (int) – the number of Monte Carlo shots

JsonToSection(jsn, vars)

Helper function that transfers a JSON file segment into GUI data structures

Parameters

jsn (dict) – the loaded JSON file
vars (dict) – the local data structure

OpenConfig(): Opens a JSON case configuration; transfers its data to the GUI

ReadFrame(f, vars)

Helper function that reads values from gridded GUI controls into the local case configuration

Parameters

f (Frame) – the GUI page to read
vars (dict) – the local data structure to update

ReadLatLong(): Updates the Latitude and Longitude from TMY3 file

ReloadFrame(f, vars)

Helper function to recreate the GUI page controls and load them with values

Parameters

f (Frame) – the GUI page to reload
vars (dict) – the section of case configuration with values to be loaded

SaveConfig(): Updates the local case configuration from the GUI, queries user for a file name, and then saves case configuration to that file

SizeMonteCarlo(n)

Initializes the Monte Carlo data structures with variable choices and samples

Parameters: n (int) – the number of Monte Carlo shots

SizeMonteCarloFrame(f)

Update the Monte Carlo page to match the number of shots and variables

Parameters: f (Frame) – the Monte Carlo GUI page

UpdateEMS(event)

UpdateMonteCarloFrame(): Transfer data from the Monte Carlo page into the case configuration

mcBand(var)

Find the band size corresponding to each Monte Carlo variable choice

Parameters: var (str) – one of ElectricCoolingParticipation, ThermostatRampMid, ThermostatOffsetLimitMid, WeekdayEveningStartMid or WeekdayEveningSetMid

mcSample(var)

Return an appropriate random value for each Monte Carlo variable choice

Parameters: var (str) – one of ElectricCoolingParticipation, ThermostatRampMid, ThermostatOffsetLimitMid, WeekdayEveningStartMid or WeekdayEveningSetMid

update_entry(ctl, val)

tesp_support.tesp_config.show_tesp_config(): Runs the GUI. Reads and writes JSON case configuration files.

tesp_support.tesp_monitor module

Presents a GUI to launch a TESP simulation and monitor its progress

Public Functions:

show_tesp_monitor: Initializes and runs the monitor GUI

References

Graphical User Interfaces with Tk

Matplotlib Animation

class tesp_support.tesp_monitor.TespMonitorGUI(master)

Bases: object

Manages a GUI with 4 plotted variables, and buttons to stop TESP

The GUI reads a JSON file with scripted shell commands to launch other FNCS federates, and a YAML file with FNCS subscriptions to update the solution status. Both JSON and YAML files are written by tesp.tesp_config The plotted variables provide a sign-of-life and sign-of-stability indication for each of the major federates in the te30 or sgip1 examples, namely GridLAB-D, PYPOWER, EnergPlus, and the substation_loop that manages a simple_auction with multiple hvac agents. If a solution appears to be unstable or must be stopped for any other reason, exiting the solution monitor will do so.

The plots are created and updated with animated and bit-blitted Matplotlib graphs hosted on a TkInter GUI. When the JSON and YAML files are loaded, the x axis is laid out to match the total TESP simulation time range.

Parameters: master –

root

the TCL Tk toolkit instance

Type: Tk

top

the top-level TCL Tk Window

Type: Window

labelvar

used to display the monitor JSON configuration file path

Type: StringVar

hrs

x-axis data array for time in hours, shared by all plots

Type: [float]

y0

y-axis data array for PYPOWER bus voltage

Type: [float]

y1

y-axis data array for EnergyPlus load

Type: [float]

y2lmp

y-axis data array for PYPOWER LMP

Type: [float]

y2auc

y-axis data array for simple_auction cleared_price

Type: [float]

y3fncs

y-axis data array for GridLAB-D load via FNCS

Type: [float]

y3gld

y-axis data array for sample-and-hold GridLAB-D load

Type: [float]

gld_load

the most recent load published by GridLAB-D; due to the deadband, this value isn’t necessary published at every FNCS time step

Type: float

y0min

the first y axis minimum value

Type: float

y0max

the first y axis maximum value

Type: float

y1min

the second y axis minimum value

Type: float

y1max

the second y axis maximum value

Type: float

y2min

the third y axis minimum value

Type: float

y2max

the third y axis maximum value

Type: float

y3min

the fourth y axis minimum value

Type: float

y3max

the fourth y axis maximum value

Type: float

hour_stop

the maximum x axis time value to plot

Type: float

ln0

the plotted PYPOWER bus voltage, color GREEN

Type: Line2D

ln1

the plotted EnergyPlus load, color RED

Type: Line2D

ln2lmp

the plotted PYPOWER locational marginal price (LMP), color BLUE

Type: Line2D

ln2auc

the plotted simple_auction cleared_price, color BLACK

Type: Line2D

ln3gld

the plotted sample-and-hold GridLAB-D substation load, color MAGENTA

Type: Line2D

ln3fncs

the plotted GridLAB-D substation load published via FNCS; may be zero if not published for the current animation frame, color CYAN

Type: Line2D

fig

animated Matplotlib figure hosted on the GUI

Type: Figure

ax

set of 4 xy axes to plot on

Type: Axes

canvas

a TCL Tk canvas that can host Matplotlib

Type: FigureCanvasTkAgg

bFNCSactive

True if a TESP simulation is running with other FNCS federates, False if not

Type: Boolean

OpenConfig(): Read the JSON configuration file for this monitor, and initialize the plot axes

Quit(): Shuts down this monitor, and also shuts down FNCS if active

expand_limits(v, vmin, vmax)

Whenever a variable meets a vertical axis limit, expand the limits with 10% padding above and below

Parameters

v (float) – the out of range value
vmin (float) – the current minimum vertical axis value
vmax (float) – the current maximum vertical axis value

Returns

the new vmin and vmax

Return type

float, float

kill_all(): Shut down all FNCS federates in TESP, except for this monitor

launch_all(): Launches the simulators, initializes FNCS and starts the animated plots

on_closing()

Verify whether the user wants to stop TESP simulations before exiting the monitor

This monitor is itself a FNCS federate, so it can not be shut down without shutting down all other FNCS federates in the TESP simulation.

update_plots(i)

This function is called by Matplotlib for each animation frame

Each time called, collect FNCS messages until the next time to plot has been reached. Then update the plot quantities and return the Line2D objects that have been updated for plotting. Check for new data outside the plotted vertical range, which triggers a full re-draw of the axes. On the last frame, finalize FNCS.

Parameters: i (int) – the animation frame number

tesp_support.tesp_monitor.show_tesp_monitor(): Creates and displays the monitor GUI

tesp_support.tesp_monitor_ercot module

Presents a GUI to launch a TESP simulation and monitor its progress

This version differs from the one in tesp_monitor, in that the user can select the FNCS federate and topic to plot. The number of monitored plots is still fixed at 4.

Public Functions:

show_tesp_monitor: Initializes and runs the monitor GUI

References

Graphical User Interfaces with Tk

Matplotlib Animation

class tesp_support.tesp_monitor_ercot.ChoosablePlot(master, color='red', xLabel='', topicDict={}, **kwargs)

Bases: Frame

Hosts one Matplotlib animation with a selected variable to plot

Parameters

master –
color (str) – Matplotlib color of the line to plot
xLabel (str) – label for the x axis
topicDict (dict) – dictionary of FNCS topic choices to plot
kwargs – arbitrary keyword arguments

topicDict

listOfTopics

root

the TCL Tk toolkit instance

Type: Tk

fig

animated Matplotlib figure hosted on the GUI

Type: Figure

ax

set of 4 xy axes to plot on

Type: Axes

xLabel

horizontal axis label

Type: str

yLabel

vertical axis label

Type: str

hrs

y

ymin

Type: float

ymax

Type: float

color

title

Type: str

ln

Type: Line2D

topicSelectionRow

topicLabel

combobox

voltageLabel

voltageBase

voltageBaseTextbox

onTopicSelected(event)

Change the GUI labels to match selected plot variable

Parameters: event –

class tesp_support.tesp_monitor_ercot.TespMonitorGUI(master)

Bases: object

Version of the monitor GUI that hosts 4 choosable plots

Parameters: master –

root

the TCL Tk toolkit instance

Type: Tk

top

the top-level TCL Tk Window

Type: Window

labelvar

used to display the monitor JSON configuration file path

Type: StringVar

plot0

first plot

Type: ChoosablePlot

plot1

second plot

Type: ChoosablePlot

plot2

third plot

Type: ChoosablePlot

plot3

fourth plot

Type: ChoosablePlot

topicDict

plots

Type: Frame

scrollbar

Type: Scrollbar

frameInCanvas

Type: Frame

canvas

a TCL Tk canvas that can host Matplotlib

Type: FigureCanvasTkAgg

bFNCSactive

True if a TESP simulation is running with other FNCS federates, False if not

Type: Boolean

CalculateValue(topic, value, plot)

Parses a value from FNCS to plot

Parameters

topic (str) – the FNCS topic
value (str) – the FNCS value
plot (ChoosablePlot) – the plot that will be updated; contains the voltage base if needed

Returns

the parsed value

Return type

float

OpenConfig(): Read the JSON configuration file for this monitor, and initialize the plot axes

Quit(): Shuts down this monitor, and also shuts down FNCS if active

kill_all(): Shut down all FNCS federates in TESP, except for this monitor

launch_all(): Launches the simulators, initializes FNCS and starts the animated plots

onFrameConfigure(event): Reset the scroll region to encompass the inner frame

on_closing()

Verify whether the user wants to stop TESP simulations before exiting the monitor

This monitor is itself a FNCS federate, so it can not be shut down without shutting down all other FNCS federates in the TESP simulation.

update_plots(i)

This function is called by Matplotlib for each animation frame

Each time called, collect FNCS messages until the next time to plot has been reached. Then update the plot quantities and return the Line2D objects that have been updated for plotting. Check for new data outside the plotted vertical range, which triggers a full re-draw of the axes. On the last frame, finalize FNCS.

Parameters: i (int) – the animation frame number

tesp_support.tesp_monitor_ercot.show_tesp_monitor(): Creates and displays the monitor GUI

tesp_support.weatherAgent module

Weather Agent

This weather agent needs an WEATHER_CONFIG environment variable to be set, which is a json file.

tesp_support.weatherAgent.convertTimeToSeconds(time)

Convert time string with unit to integer in seconds

It only parse unit in day, hour, minute and second. It will not recognize week, month, year, millisecond, microsecond or nanosecond, they can be added if needed.

Parameters: time – str time with unit
Returns: int represent the input time in second

tesp_support.weatherAgent.deltaTimeToResmapleFreq(time)

Convert time unit to a resampling frequency that can be recognized by pandas.DataFrame.resample()

It only parse unit in day, hour, minute and second. It won’t recognize week, month, year, millisecond, microsecond or nanosecond, they can be added if needed.

Parameters: time – str time with unit
Returns: str time with resample frequency

tesp_support.weatherAgent.findDeltaTimeMultiplier(time)

find the multiplier to convert delta_time to seconds

It only parse unit in day, hour, minute and second. It won’t recognize week, month, year, millisecond, microsecond or nanosecond, they can be added if needed.

Parameters: time – str time with unit
Returns: int the multiplier to convert delta_time to seconds

tesp_support.weatherAgent.show_resource_consumption()

tesp_support.weatherAgent.startWeatherAgent(file)

the weather agent publishes weather data as configured by the json file

Parameters: file – string the weather data file
Returns: nothing

tesp_support.weatherAgent.usage()

class tesp_support.weatherAgent.weather_forecast(variable, period, W_dict)

Bases: object

This object includes the error to a weather variable

Parameters

variable (str) – Type of weather variable being forecasted
period (int) – period of the sinusoidal bias
W_dict (dict) – dictionary for specifying the generation of the error envelope

weather_variable

Type of weather variable being forecasted

Type: str

# Type of error insertion

distribution

type of distribution –> 0 uniform;1 triangular;2 truncated normal the standard deviation is computed for 95% of values to be within bounds in a conventional normal distribution

Type: int

P_e_bias

pu maximum bias at first hour –> [0 to 1]

Type: float

P_e_envelope

pu maximum error from mean values –> [0 to 1]

Type: float

Lower_e_bound

pu of the maximum error at the first hour –> [0 to 1]

Type: float

# Bias variable

biasM

sinusoidal bias for altering the error envelope

Type: float) (1 X period

Period_bias

period of the sinusoidal bias

Type: int

get_truncated_normal(EL, EH): Truncated normal distribution

make_forecast(weather, t=0)

Include error to a known weather variable

Parameters

weather (float) (1 x desired number of hours ahead) – known weather variable
t (int) – time in hours

Returns

weather variable with included error ENV_U (float) (1 x desired number of hours ahead): envelope with bias upper bound ENV_l (float) (1 x desired number of hours ahead): envelope with bias lower bound

Return type

weather_f (float) (1 x desired number of hours ahead)