3. Data requirements

Currently, network generation and hazard generation are not automated in the integrated simulation model. Therefore, in order to use the model, three types of data are to be manually fed into the model, namely the water, power and transportation networks, interdependency data, and infrastructure disruption data. There datasets need to be in specific formats which are compatible with the model. In this chapter, the details such as file formats and information to be included in the input files are discussed in detail.

3.1. Infrastructre Networks

The infrastructure network data must be compatible with the respective infrastructure model packages enlisted in Table 1. In addition, disruptions to certain components belonging to the infrastructure systems are not supported as of now. The individual networks must be constructed taking into the above aspects into consideration.

3.1.1. Water distribution system

Since the integrated simulation model handles the water distribution network models using wntr package, the input file must be in .inp format. Some examples of water network files can be found in wntr Github repository1. The water network can be either built using EPANET or wntr package. Currently, the water network simulation is performed using WNTRSimulator in the wntr package. Therefore, there are certain exceptions which must be considered while generating the water network file. These exceptions can be found in wntr’s software framework and limitations page.

In addition, the integrated simulation model identifies the component details such as infrastructure type, component type, etc. using the component names. The name of the components must follow the nomenclature. The details of nomenclature and whether a component can be failed in the model is presented in Table 3.1. For example, a water pump can be named as W_WP1. Integers must be followed by the prefix to name components belonging to same category.

Table 3.1. Nomenclature for water network components

Prefix	Component name	Disruption supported	Prefix	Component name	Disruption supported
W_WP	Pump	Yes	W_PMA	Main Pipe	Yes
W_R	Reservoir	No	W_PHC	Hydrant Connection Pipe	Yes
W_P	Pipe	Yes	W_PV	Valve converted to Pipe	Yes
W_PSC	Service Connection Pipe	Yes	W_J	Junction	No
W_T	Tank	Yes

3.1.2. Power systems

The power system is modeled using pandapower package, and therefore the network file must be in *.json format. Several tutorials for generating power system networks using pandapower are available in the package’s tutorial page. Similar to water network components, naming of power system components must also follow the stipulated nomenclature presented in Table 3. Currently, power networks are modeled as three-phase systems. Therefore, single-phase components in pandapower are not supported. Integers must be followed by the prefix to name components belonging to same category.

Table 3.2. Nomenclature for power network components

Prefix	Component name	Disruption supported	Prefix	Component name	Disruption supported
P_B	Bus	Yes	P_L	Line	Yes
P_BLO	Bus connected to load	Yes	P_TF	Transformer	Yes
P_BS	Bus connected to switch	Yes	P_TFEG	Transformer connected to external grid	Yes
P_BEG	Bus connected to external grid connection	yes	P_S	Switch	Yes
P_LO	Load	Yes	P_SEG	Switch connected to external grid	Yes
P_MP	Motor	Yes	P_SLI	Switch connected to Line	Yes
P_G	Generator	No	P_EG	External Grid connection	Yes

3.1.3. Transportation system

The transportation system is simulated using the static traffic assignment model developed by Prof. Stephen Boyles of Department of Civil, Architectural and Environmental Engineering, The University of Texas at Austin. The transportation network data must be in the TNTP data format. More details on the format and example networks can be found in Ben Stabler’s Github page4. The naming of the transportation network components must follow the nomenclature presented in Table 3.3. Integers must be followed by the prefix to name components belonging to same category.

Table 3.3. Nomenclature for transportation network components

Prefix	Component name	Disruption supported
T_J	Junction	No
T_L	Link	Yes

3.2. Infrastructure dependencies

Data related to water-power dependencies between infrastructure components must be provided separately in a .csv file. The file must contain water_id and power_id fields which represent the water- and power component names (Table 3.4). The model will determine the type of the waterand power components and construct the dependencies accordingly.

Table 3.4. Format of disruption data

water_id	power_id
W_WP9	P_MP1
W_R9	P_G3

3.3. Infrastructure disruption data

The third data input is related to disrupted components. Similar to the dependency input file, the disruption data also needs to be provided in *.dat format. The file should have three fields, namely time_stamp (time of disruption in seconds), components (name of the component that is included in any of the three networks), fail_perc (percentage of damage), and recovery_time (recovery time in hours). An example for the dependency data is presented in Table 3.5.

table 3.5. Format of disruption data

time_stamp	components	fail_perc	recovery_time
7200	W_PMA1	50	7
7200	W_MP2	75	138
7200	T_L438	25	13

While the percentage of damage of component is not used in the current model, this may be incorporated in the model to find the repair duration estimate in the future versions.