Parsing CIM XML Network Data¶

This tutorial demonstrates how to parse a CIM (Common Information Model) XML file from UK Power Networks' LTDS publication and build a pandapower network from it.

What is CIM?¶

The Common Information Model (CIM) is an IEC standard (IEC 61970/61968) used by Distribution Network Operators to describe power system equipment, topology, and connectivity in a vendor-neutral XML/RDF format.

UKPN publishes LTDS network models as CIM XML files. Each file contains:

CIM Class	Description
`Substation`	Named substations (Grid, Primary, etc.)
`VoltageLevel`	Voltage levels within a substation
`PowerTransformer` / `PowerTransformerEnd`	Transformers with impedances and ratings
`ACLineSegment`	Overhead lines and cables with R, X, B parameters
`EnergyConsumer`	Load points
`ConnectivityNode` / `Terminal`	Topology and connectivity
`PhotoVoltaicUnit`, `BatteryUnit`, `SynchronousMachine`	Generation and storage

How to Get CIM Data¶

The LTDS CIM dataset is a "Shared" dataset that requires special access. To request access:

Register and login to the UKPN Open Data Portal
Visit the LTDS CIM page and complete the Shared Data Request Form

Once approved, CIM data is published as XML file attachments (one per licence area: EPN, SPN, LPN). You can download the XML files directly from the portal.

This tutorial includes a small example file so you can follow along without needing portal access.

Prerequisites¶

Complete 01-getting-started.ipynb first
Install pandapower: pip install "ukpyn[all]"

The example file used here is cim-example.xml, an excerpt from the EPN licence area LTDS Equipment profile.

1. Load and Explore the CIM XML¶

In [ ]:

Copied!





import xml.etree.ElementTree as ET
from collections import Counter
from pathlib import Path

CIM_FILE = Path("cim-example.xml")

tree = ET.parse(CIM_FILE)
root = tree.getroot()

# CIM namespaces used in UKPN LTDS files
NS = {
    "cim": "http://iec.ch/TC57/CIM100#",
    "rdf": "http://www.w3.org/1999/02/22-rdf-syntax-ns#",
    "md": "http://iec.ch/TC57/61970-552/ModelDescription/1#",
    "gb": "http://ofgem.gov.uk/ns/CIM/LTDS/Extensions#",
    "eu": "http://iec.ch/TC57/CIM100-European#",
}

RDF_ID = "{http://www.w3.org/1999/02/22-rdf-syntax-ns#}ID"
RDF_RES = "{http://www.w3.org/1999/02/22-rdf-syntax-ns#}resource"

# Count element types
counter = Counter()
for elem in root:
    tag = elem.tag.split("}")[-1] if "}" in elem.tag else elem.tag
    counter[tag] += 1

print(f"Total CIM elements: {sum(counter.values()):,}")
print(f"Distinct types: {len(counter)}\n")
for tag, count in counter.most_common(15):
    print(f"  {tag:35s} {count:>6,}")
import xml.etree.ElementTree as ET
from collections import Counter
from pathlib import Path

CIM_FILE = Path("cim-example.xml")

tree = ET.parse(CIM_FILE)
root = tree.getroot()

# CIM namespaces used in UKPN LTDS files
NS = {
    "cim": "http://iec.ch/TC57/CIM100#",
    "rdf": "http://www.w3.org/1999/02/22-rdf-syntax-ns#",
    "md": "http://iec.ch/TC57/61970-552/ModelDescription/1#",
    "gb": "http://ofgem.gov.uk/ns/CIM/LTDS/Extensions#",
    "eu": "http://iec.ch/TC57/CIM100-European#",
}

RDF_ID = "{http://www.w3.org/1999/02/22-rdf-syntax-ns#}ID"
RDF_RES = "{http://www.w3.org/1999/02/22-rdf-syntax-ns#}resource"

# Count element types
counter = Counter()
for elem in root:
    tag = elem.tag.split("}")[-1] if "}" in elem.tag else elem.tag
    counter[tag] += 1

print(f"Total CIM elements: {sum(counter.values()):,}")
print(f"Distinct types: {len(counter)}\n")
for tag, count in counter.most_common(15):
    print(f"  {tag:35s} {count:>6,}")

2. Extract Key Network Components¶

We'll build lookup dictionaries for the main CIM classes. Each element's rdf:ID attribute is the unique key used to cross-reference related objects.

In [ ]:

Copied!





def get_text(elem, child_tag):
    """Get text content of a child element, or None."""
    child = elem.find(child_tag, NS)
    return child.text if child is not None and child.text else None


def get_ref(elem, child_tag):
    """Get rdf:resource reference (stripped of leading '#' and '_'), or None."""
    child = elem.find(child_tag, NS)
    if child is not None:
        ref = child.get(RDF_RES, "")
        return ref.lstrip("#_") if ref else None
    return None


def get_float(elem, child_tag):
    """Get float value of a child element, or None."""
    val = get_text(elem, child_tag)
    return float(val) if val is not None else None
def get_text(elem, child_tag):
    """Get text content of a child element, or None."""
    child = elem.find(child_tag, NS)
    return child.text if child is not None and child.text else None


def get_ref(elem, child_tag):
    """Get rdf:resource reference (stripped of leading '#' and '_'), or None."""
    child = elem.find(child_tag, NS)
    if child is not None:
        ref = child.get(RDF_RES, "")
        return ref.lstrip("#_") if ref else None
    return None


def get_float(elem, child_tag):
    """Get float value of a child element, or None."""
    val = get_text(elem, child_tag)
    return float(val) if val is not None else None

In [ ]:

Copied!





# --- BaseVoltage lookup (ID -> kV) ---
base_voltages = {}
for bv in root.findall("cim:BaseVoltage", NS):
    bv_id = bv.get(RDF_ID).lstrip("_")
    kv = get_float(bv, "cim:BaseVoltage.nominalVoltage")
    base_voltages[bv_id] = kv

print(f"Base voltages: {len(base_voltages)}")
for _bv_id, kv in sorted(base_voltages.items(), key=lambda x: x[1] or 0):
    print(f"  {kv} kV")
# --- BaseVoltage lookup (ID -> kV) ---
base_voltages = {}
for bv in root.findall("cim:BaseVoltage", NS):
    bv_id = bv.get(RDF_ID).lstrip("_")
    kv = get_float(bv, "cim:BaseVoltage.nominalVoltage")
    base_voltages[bv_id] = kv

print(f"Base voltages: {len(base_voltages)}")
for _bv_id, kv in sorted(base_voltages.items(), key=lambda x: x[1] or 0):
    print(f"  {kv} kV")

In [ ]:

Copied!





import pandas as pd

# --- Substations ---
substations = {}
for sub in root.findall("cim:Substation", NS):
    sub_id = sub.get(RDF_ID).lstrip("_")
    substations[sub_id] = {
        "name": get_text(sub, "cim:IdentifiedObject.name"),
        "region": get_ref(sub, "cim:Substation.Region"),
    }

print(f"Substations: {len(substations)}")
df_subs = pd.DataFrame.from_dict(substations, orient="index")
df_subs.head(10)
import pandas as pd

# --- Substations ---
substations = {}
for sub in root.findall("cim:Substation", NS):
    sub_id = sub.get(RDF_ID).lstrip("_")
    substations[sub_id] = {
        "name": get_text(sub, "cim:IdentifiedObject.name"),
        "region": get_ref(sub, "cim:Substation.Region"),
    }

print(f"Substations: {len(substations)}")
df_subs = pd.DataFrame.from_dict(substations, orient="index")
df_subs.head(10)

In [ ]:

Copied!





# --- Voltage Levels ---
voltage_levels = {}
for vl in root.findall("cim:VoltageLevel", NS):
    vl_id = vl.get(RDF_ID).lstrip("_")
    bv_ref = get_ref(vl, "cim:VoltageLevel.BaseVoltage")
    voltage_levels[vl_id] = {
        "name": get_text(vl, "cim:IdentifiedObject.name"),
        "substation_id": get_ref(vl, "cim:VoltageLevel.Substation"),
        "base_kv": base_voltages.get(bv_ref) if bv_ref else None,
    }

print(f"Voltage levels: {len(voltage_levels)}")
df_vl = pd.DataFrame.from_dict(voltage_levels, orient="index")
df_vl["base_kv"].value_counts().sort_index()
# --- Voltage Levels ---
voltage_levels = {}
for vl in root.findall("cim:VoltageLevel", NS):
    vl_id = vl.get(RDF_ID).lstrip("_")
    bv_ref = get_ref(vl, "cim:VoltageLevel.BaseVoltage")
    voltage_levels[vl_id] = {
        "name": get_text(vl, "cim:IdentifiedObject.name"),
        "substation_id": get_ref(vl, "cim:VoltageLevel.Substation"),
        "base_kv": base_voltages.get(bv_ref) if bv_ref else None,
    }

print(f"Voltage levels: {len(voltage_levels)}")
df_vl = pd.DataFrame.from_dict(voltage_levels, orient="index")
df_vl["base_kv"].value_counts().sort_index()

3. Parse Transformers¶

CIM splits transformer data across two classes:

PowerTransformer — the transformer itself (container, name)
PowerTransformerEnd — each winding's electrical parameters (ratedS, ratedU, r, x)

We join them to get a complete picture.

In [ ]:

Copied!





# --- PowerTransformers ---
transformers = {}
for pt in root.findall("cim:PowerTransformer", NS):
    pt_id = pt.get(RDF_ID).lstrip("_")
    transformers[pt_id] = {
        "name": get_text(pt, "cim:IdentifiedObject.name"),
        "container": get_ref(pt, "cim:Equipment.EquipmentContainer"),
        "ends": [],  # populated below
    }

# --- PowerTransformerEnds ---
for pte in root.findall("cim:PowerTransformerEnd", NS):
    pt_ref = get_ref(pte, "cim:PowerTransformerEnd.PowerTransformer")
    if pt_ref and pt_ref in transformers:
        bv_ref = get_ref(pte, "cim:TransformerEnd.BaseVoltage")
        transformers[pt_ref]["ends"].append(
            {
                "end_number": int(get_text(pte, "cim:TransformerEnd.endNumber") or 0),
                "ratedS_mva": get_float(pte, "cim:PowerTransformerEnd.ratedS"),
                "ratedU_kv": get_float(pte, "cim:PowerTransformerEnd.ratedU"),
                "r_ohm": get_float(pte, "cim:PowerTransformerEnd.r"),
                "x_ohm": get_float(pte, "cim:PowerTransformerEnd.x"),
                "base_kv": base_voltages.get(bv_ref) if bv_ref else None,
            }
        )

print(f"Transformers: {len(transformers)}")
print(f"Transformer ends: {sum(len(t['ends']) for t in transformers.values())}")

# Show a sample
sample_id = next(k for k, v in transformers.items() if len(v["ends"]) == 2)
sample = transformers[sample_id]
print(f"\nSample: {sample['name']}")
for end in sorted(sample["ends"], key=lambda e: e["end_number"]):
    print(
        f"  End {end['end_number']}: {end['ratedU_kv']} kV, {end['ratedS_mva']} MVA, "
        f"R={end['r_ohm']}, X={end['x_ohm']}"
    )
# --- PowerTransformers ---
transformers = {}
for pt in root.findall("cim:PowerTransformer", NS):
    pt_id = pt.get(RDF_ID).lstrip("_")
    transformers[pt_id] = {
        "name": get_text(pt, "cim:IdentifiedObject.name"),
        "container": get_ref(pt, "cim:Equipment.EquipmentContainer"),
        "ends": [],  # populated below
    }

# --- PowerTransformerEnds ---
for pte in root.findall("cim:PowerTransformerEnd", NS):
    pt_ref = get_ref(pte, "cim:PowerTransformerEnd.PowerTransformer")
    if pt_ref and pt_ref in transformers:
        bv_ref = get_ref(pte, "cim:TransformerEnd.BaseVoltage")
        transformers[pt_ref]["ends"].append(
            {
                "end_number": int(get_text(pte, "cim:TransformerEnd.endNumber") or 0),
                "ratedS_mva": get_float(pte, "cim:PowerTransformerEnd.ratedS"),
                "ratedU_kv": get_float(pte, "cim:PowerTransformerEnd.ratedU"),
                "r_ohm": get_float(pte, "cim:PowerTransformerEnd.r"),
                "x_ohm": get_float(pte, "cim:PowerTransformerEnd.x"),
                "base_kv": base_voltages.get(bv_ref) if bv_ref else None,
            }
        )

print(f"Transformers: {len(transformers)}")
print(f"Transformer ends: {sum(len(t['ends']) for t in transformers.values())}")

# Show a sample
sample_id = next(k for k, v in transformers.items() if len(v["ends"]) == 2)
sample = transformers[sample_id]
print(f"\nSample: {sample['name']}")
for end in sorted(sample["ends"], key=lambda e: e["end_number"]):
    print(
        f"  End {end['end_number']}: {end['ratedU_kv']} kV, {end['ratedS_mva']} MVA, "
        f"R={end['r_ohm']}, X={end['x_ohm']}"
    )

4. Parse AC Line Segments¶

In [ ]:

Copied!





lines = []
for seg in root.findall("cim:ACLineSegment", NS):
    bv_ref = get_ref(seg, "cim:ConductingEquipment.BaseVoltage")
    lines.append(
        {
            "id": seg.get(RDF_ID).lstrip("_"),
            "name": get_text(seg, "cim:IdentifiedObject.name"),
            "r_ohm": get_float(seg, "cim:ACLineSegment.r"),
            "x_ohm": get_float(seg, "cim:ACLineSegment.x"),
            "bch_s": get_float(seg, "cim:ACLineSegment.bch"),
            "length_km": get_float(seg, "cim:Conductor.length"),
            "base_kv": base_voltages.get(bv_ref) if bv_ref else None,
        }
    )

df_lines = pd.DataFrame(lines)
print(f"AC Line Segments: {len(df_lines)}")
df_lines.head(10)
lines = []
for seg in root.findall("cim:ACLineSegment", NS):
    bv_ref = get_ref(seg, "cim:ConductingEquipment.BaseVoltage")
    lines.append(
        {
            "id": seg.get(RDF_ID).lstrip("_"),
            "name": get_text(seg, "cim:IdentifiedObject.name"),
            "r_ohm": get_float(seg, "cim:ACLineSegment.r"),
            "x_ohm": get_float(seg, "cim:ACLineSegment.x"),
            "bch_s": get_float(seg, "cim:ACLineSegment.bch"),
            "length_km": get_float(seg, "cim:Conductor.length"),
            "base_kv": base_voltages.get(bv_ref) if bv_ref else None,
        }
    )

df_lines = pd.DataFrame(lines)
print(f"AC Line Segments: {len(df_lines)}")
df_lines.head(10)

In [ ]:

Copied!

# Voltage distribution of lines
print("Lines by voltage level:")
df_lines["base_kv"].value_counts().sort_index()
# Voltage distribution of lines
print("Lines by voltage level:")
df_lines["base_kv"].value_counts().sort_index()

5. Parse Loads (EnergyConsumers)¶

In [ ]:

Copied!





loads = []
for ec in root.findall("cim:EnergyConsumer", NS):
    loads.append(
        {
            "id": ec.get(RDF_ID).lstrip("_"),
            "name": get_text(ec, "cim:IdentifiedObject.name"),
            "description": get_text(ec, "cim:IdentifiedObject.description"),
            "container": get_ref(ec, "cim:Equipment.EquipmentContainer"),
        }
    )

df_loads = pd.DataFrame(loads)
print(f"Energy Consumers (loads): {len(df_loads)}")
df_loads.head(10)
loads = []
for ec in root.findall("cim:EnergyConsumer", NS):
    loads.append(
        {
            "id": ec.get(RDF_ID).lstrip("_"),
            "name": get_text(ec, "cim:IdentifiedObject.name"),
            "description": get_text(ec, "cim:IdentifiedObject.description"),
            "container": get_ref(ec, "cim:Equipment.EquipmentContainer"),
        }
    )

df_loads = pd.DataFrame(loads)
print(f"Energy Consumers (loads): {len(df_loads)}")
df_loads.head(10)

6. Parse Generation and Storage¶

CIM represents generation with several classes:

SynchronousMachine — conventional generators
PhotoVoltaicUnit — solar PV
BatteryUnit — battery energy storage
WindGeneratingUnit — wind turbines

In [ ]:

Copied!





# Solar PV
pv_units = []
for pv in root.findall("cim:PhotoVoltaicUnit", NS):
    pv_units.append(
        {
            "name": get_text(pv, "cim:IdentifiedObject.name"),
            "maxP_mw": get_float(pv, "cim:PowerElectronicsUnit.maxP"),
        }
    )

# Battery storage
batteries = []
for bat in root.findall("cim:BatteryUnit", NS):
    batteries.append(
        {
            "name": get_text(bat, "cim:IdentifiedObject.name"),
            "maxP_mw": get_float(bat, "cim:PowerElectronicsUnit.maxP"),
            "ratedE_mwh": get_float(bat, "cim:BatteryUnit.ratedE"),
        }
    )

# Synchronous machines
sync_machines = []
for sm in root.findall("cim:SynchronousMachine", NS):
    sync_machines.append(
        {
            "name": get_text(sm, "cim:IdentifiedObject.name"),
            "maxExportP_mw": get_float(sm, "cim:RotatingMachine.maxExportP"),
        }
    )

# Wind
wind_units = []
for wu in root.findall("cim:WindGeneratingUnit", NS):
    wind_units.append(
        {
            "name": get_text(wu, "cim:IdentifiedObject.name"),
            "maxP_mw": get_float(wu, "cim:GeneratingUnit.maxOperatingP"),
        }
    )

print(
    f"Solar PV units:        {len(pv_units):>4}  — Total {sum(p['maxP_mw'] or 0 for p in pv_units):.1f} MW"
)
print(
    f"Battery units:         {len(batteries):>4}  — Total {sum(b['maxP_mw'] or 0 for b in batteries):.1f} MW"
)
print(
    f"Synchronous machines:  {len(sync_machines):>4}  — Total {sum(s['maxExportP_mw'] or 0 for s in sync_machines):.1f} MW"
)
print(f"Wind units:            {len(wind_units):>4}")
# Solar PV
pv_units = []
for pv in root.findall("cim:PhotoVoltaicUnit", NS):
    pv_units.append(
        {
            "name": get_text(pv, "cim:IdentifiedObject.name"),
            "maxP_mw": get_float(pv, "cim:PowerElectronicsUnit.maxP"),
        }
    )

# Battery storage
batteries = []
for bat in root.findall("cim:BatteryUnit", NS):
    batteries.append(
        {
            "name": get_text(bat, "cim:IdentifiedObject.name"),
            "maxP_mw": get_float(bat, "cim:PowerElectronicsUnit.maxP"),
            "ratedE_mwh": get_float(bat, "cim:BatteryUnit.ratedE"),
        }
    )

# Synchronous machines
sync_machines = []
for sm in root.findall("cim:SynchronousMachine", NS):
    sync_machines.append(
        {
            "name": get_text(sm, "cim:IdentifiedObject.name"),
            "maxExportP_mw": get_float(sm, "cim:RotatingMachine.maxExportP"),
        }
    )

# Wind
wind_units = []
for wu in root.findall("cim:WindGeneratingUnit", NS):
    wind_units.append(
        {
            "name": get_text(wu, "cim:IdentifiedObject.name"),
            "maxP_mw": get_float(wu, "cim:GeneratingUnit.maxOperatingP"),
        }
    )

print(
    f"Solar PV units:        {len(pv_units):>4}  — Total {sum(p['maxP_mw'] or 0 for p in pv_units):.1f} MW"
)
print(
    f"Battery units:         {len(batteries):>4}  — Total {sum(b['maxP_mw'] or 0 for b in batteries):.1f} MW"
)
print(
    f"Synchronous machines:  {len(sync_machines):>4}  — Total {sum(s['maxExportP_mw'] or 0 for s in sync_machines):.1f} MW"
)
print(f"Wind units:            {len(wind_units):>4}")

7. Build a pandapower Network from CIM Data¶

We'll select a single substation with 33/11 kV transformers and build a small pandapower model from the CIM data. This mirrors the approach in 09-pandapower-import.ipynb but sourced directly from XML.

In [ ]:

Copied!





# Find substations that have 2-winding transformers with both HV and LV data
# First, map EquipmentContainer (substation/voltage-level) -> substation name
container_to_sub = {}
for vl_id, vl in voltage_levels.items():
    sub_id = vl["substation_id"]
    if sub_id and sub_id in substations:
        container_to_sub[vl_id] = substations[sub_id]["name"]
for sub_id, sub in substations.items():
    container_to_sub[sub_id] = sub["name"]

# Find transformers with exactly 2 ends and real impedance data
good_trafos = []
for pt_id, pt in transformers.items():
    ends = pt["ends"]
    if len(ends) != 2:
        continue
    ends_sorted = sorted(ends, key=lambda e: e["end_number"])
    hv_end = ends_sorted[0]  # end 1 = HV
    lv_end = ends_sorted[1]  # end 2 = LV
    if hv_end["ratedS_mva"] and hv_end["ratedU_kv"] and hv_end["x_ohm"]:
        sub_name = container_to_sub.get(pt["container"], "Unknown")
        good_trafos.append(
            {
                "id": pt_id,
                "name": pt["name"],
                "substation": sub_name,
                "hv_kv": hv_end["ratedU_kv"],
                "lv_kv": lv_end["ratedU_kv"],
                "sn_mva": hv_end["ratedS_mva"],
                "r_ohm": hv_end["r_ohm"],
                "x_ohm": hv_end["x_ohm"],
            }
        )

df_trafos = pd.DataFrame(good_trafos)
print(f"Transformers with complete 2-winding data: {len(df_trafos)}")

# Show substations with the most transformers
print("\nTop substations by transformer count:")
df_trafos["substation"].value_counts().head(10)
# Find substations that have 2-winding transformers with both HV and LV data
# First, map EquipmentContainer (substation/voltage-level) -> substation name
container_to_sub = {}
for vl_id, vl in voltage_levels.items():
    sub_id = vl["substation_id"]
    if sub_id and sub_id in substations:
        container_to_sub[vl_id] = substations[sub_id]["name"]
for sub_id, sub in substations.items():
    container_to_sub[sub_id] = sub["name"]

# Find transformers with exactly 2 ends and real impedance data
good_trafos = []
for pt_id, pt in transformers.items():
    ends = pt["ends"]
    if len(ends) != 2:
        continue
    ends_sorted = sorted(ends, key=lambda e: e["end_number"])
    hv_end = ends_sorted[0]  # end 1 = HV
    lv_end = ends_sorted[1]  # end 2 = LV
    if hv_end["ratedS_mva"] and hv_end["ratedU_kv"] and hv_end["x_ohm"]:
        sub_name = container_to_sub.get(pt["container"], "Unknown")
        good_trafos.append(
            {
                "id": pt_id,
                "name": pt["name"],
                "substation": sub_name,
                "hv_kv": hv_end["ratedU_kv"],
                "lv_kv": lv_end["ratedU_kv"],
                "sn_mva": hv_end["ratedS_mva"],
                "r_ohm": hv_end["r_ohm"],
                "x_ohm": hv_end["x_ohm"],
            }
        )

df_trafos = pd.DataFrame(good_trafos)
print(f"Transformers with complete 2-winding data: {len(df_trafos)}")

# Show substations with the most transformers
print("\nTop substations by transformer count:")
df_trafos["substation"].value_counts().head(10)

In [ ]:

Copied!





import math

import pandapower as pp

# Pick a substation — choose one with a few transformers
sub_counts = df_trafos["substation"].value_counts()
TARGET_SUB = sub_counts[(sub_counts >= 2) & (sub_counts <= 6)].index[0]
print(f"Selected substation: {TARGET_SUB}")

df_sub_trafos = df_trafos[df_trafos["substation"] == TARGET_SUB].copy()
display(df_sub_trafos[["name", "hv_kv", "lv_kv", "sn_mva", "r_ohm", "x_ohm"]])
import math

import pandapower as pp

# Pick a substation — choose one with a few transformers
sub_counts = df_trafos["substation"].value_counts()
TARGET_SUB = sub_counts[(sub_counts >= 2) & (sub_counts <= 6)].index[0]
print(f"Selected substation: {TARGET_SUB}")

df_sub_trafos = df_trafos[df_trafos["substation"] == TARGET_SUB].copy()
display(df_sub_trafos[["name", "hv_kv", "lv_kv", "sn_mva", "r_ohm", "x_ohm"]])

In [ ]:

Copied!





# Build the pandapower network
net = pp.create_empty_network(name=f"CIM — {TARGET_SUB}")

# Create buses from unique voltage levels
bus_map = {}  # (name, kv) -> bus index

for _, row in df_sub_trafos.iterrows():
    for _side, kv_col in [("HV", "hv_kv"), ("LV", "lv_kv")]:
        kv = row[kv_col]
        bus_name = f"{TARGET_SUB} {kv:.0f}kV"
        key = (bus_name, kv)
        if key not in bus_map:
            bus_map[key] = pp.create_bus(net, vn_kv=kv, name=bus_name)

# Create transformers
for _, row in df_sub_trafos.iterrows():
    hv_bus = bus_map[(f"{TARGET_SUB} {row['hv_kv']:.0f}kV", row["hv_kv"])]
    lv_bus = bus_map[(f"{TARGET_SUB} {row['lv_kv']:.0f}kV", row["lv_kv"])]
    sn_mva = row["sn_mva"]
    vn_hv = row["hv_kv"]

    # Convert ohmic impedance to percent on transformer base
    z_base = (vn_hv**2) / sn_mva
    vkr_percent = (row["r_ohm"] / z_base) * 100
    vkx_percent = (row["x_ohm"] / z_base) * 100
    vk_percent = math.sqrt(vkr_percent**2 + vkx_percent**2)

    pp.create_transformer_from_parameters(
        net,
        hv_bus=hv_bus,
        lv_bus=lv_bus,
        sn_mva=sn_mva,
        vn_hv_kv=vn_hv,
        vn_lv_kv=row["lv_kv"],
        vk_percent=vk_percent,
        vkr_percent=vkr_percent,
        pfe_kw=0,
        i0_percent=0.1,
        name=row["name"],
    )

# External grid at the HV bus
gsp_bus = net.bus["vn_kv"].idxmax()
pp.create_ext_grid(net, bus=gsp_bus, vm_pu=1.0, name="Grid")

# Add a representative load on each LV bus
for bus_idx in net.bus.index:
    if bus_idx == gsp_bus:
        continue
    pp.create_load(
        net,
        bus=bus_idx,
        p_mw=5.0,
        q_mvar=1.5,
        name=f"Load {net.bus.loc[bus_idx, 'name']}",
    )

print(f"Buses: {len(net.bus)}, Transformers: {len(net.trafo)}, Loads: {len(net.load)}")
display(net.bus)
display(
    net.trafo[["name", "sn_mva", "vn_hv_kv", "vn_lv_kv", "vk_percent", "vkr_percent"]]
)
# Build the pandapower network
net = pp.create_empty_network(name=f"CIM — {TARGET_SUB}")

# Create buses from unique voltage levels
bus_map = {}  # (name, kv) -> bus index

for _, row in df_sub_trafos.iterrows():
    for _side, kv_col in [("HV", "hv_kv"), ("LV", "lv_kv")]:
        kv = row[kv_col]
        bus_name = f"{TARGET_SUB} {kv:.0f}kV"
        key = (bus_name, kv)
        if key not in bus_map:
            bus_map[key] = pp.create_bus(net, vn_kv=kv, name=bus_name)

# Create transformers
for _, row in df_sub_trafos.iterrows():
    hv_bus = bus_map[(f"{TARGET_SUB} {row['hv_kv']:.0f}kV", row["hv_kv"])]
    lv_bus = bus_map[(f"{TARGET_SUB} {row['lv_kv']:.0f}kV", row["lv_kv"])]
    sn_mva = row["sn_mva"]
    vn_hv = row["hv_kv"]

    # Convert ohmic impedance to percent on transformer base
    z_base = (vn_hv**2) / sn_mva
    vkr_percent = (row["r_ohm"] / z_base) * 100
    vkx_percent = (row["x_ohm"] / z_base) * 100
    vk_percent = math.sqrt(vkr_percent**2 + vkx_percent**2)

    pp.create_transformer_from_parameters(
        net,
        hv_bus=hv_bus,
        lv_bus=lv_bus,
        sn_mva=sn_mva,
        vn_hv_kv=vn_hv,
        vn_lv_kv=row["lv_kv"],
        vk_percent=vk_percent,
        vkr_percent=vkr_percent,
        pfe_kw=0,
        i0_percent=0.1,
        name=row["name"],
    )

# External grid at the HV bus
gsp_bus = net.bus["vn_kv"].idxmax()
pp.create_ext_grid(net, bus=gsp_bus, vm_pu=1.0, name="Grid")

# Add a representative load on each LV bus
for bus_idx in net.bus.index:
    if bus_idx == gsp_bus:
        continue
    pp.create_load(
        net,
        bus=bus_idx,
        p_mw=5.0,
        q_mvar=1.5,
        name=f"Load {net.bus.loc[bus_idx, 'name']}",
    )

print(f"Buses: {len(net.bus)}, Transformers: {len(net.trafo)}, Loads: {len(net.load)}")
display(net.bus)
display(
    net.trafo[["name", "sn_mva", "vn_hv_kv", "vn_lv_kv", "vk_percent", "vkr_percent"]]
)

8. Run Power Flow¶

In [ ]:

Copied!





pp.runpp(net, algorithm="nr", calculate_voltage_angles=True)

print("Power flow converged!\n")
print("Bus Results:")
display(net.res_bus.join(net.bus["name"]))

print("\nTransformer Loading:")
trafo_results = net.res_trafo[["p_hv_mw", "q_hv_mvar", "loading_percent"]].copy()
trafo_results.insert(0, "name", net.trafo["name"])
display(trafo_results)
pp.runpp(net, algorithm="nr", calculate_voltage_angles=True)

print("Power flow converged!\n")
print("Bus Results:")
display(net.res_bus.join(net.bus["name"]))

print("\nTransformer Loading:")
trafo_results = net.res_trafo[["p_hv_mw", "q_hv_mvar", "loading_percent"]].copy()
trafo_results.insert(0, "name", net.trafo["name"])
display(trafo_results)

9. Network Statistics Summary¶

In [ ]:

Copied!





print("=" * 50)
print(f"CIM FILE SUMMARY — {CIM_FILE.name}")
print("=" * 50)
print(f"  Substations:          {len(substations):>6,}")
print(f"  Voltage levels:       {len(voltage_levels):>6,}")
print(f"  Transformers:         {len(transformers):>6,}")
print(f"  AC line segments:     {len(df_lines):>6,}")
print(f"  Loads:                {len(df_loads):>6,}")
print(
    f"  Solar PV:             {len(pv_units):>6,}  ({sum(p['maxP_mw'] or 0 for p in pv_units):.0f} MW)"
)
print(
    f"  Batteries:            {len(batteries):>6,}  ({sum(b['maxP_mw'] or 0 for b in batteries):.0f} MW)"
)
print(
    f"  Synchronous machines: {len(sync_machines):>6,}  ({sum(s['maxExportP_mw'] or 0 for s in sync_machines):.0f} MW)"
)
print(f"  Wind units:           {len(wind_units):>6,}")
print("=" * 50)
print(f"CIM FILE SUMMARY — {CIM_FILE.name}")
print("=" * 50)
print(f"  Substations:          {len(substations):>6,}")
print(f"  Voltage levels:       {len(voltage_levels):>6,}")
print(f"  Transformers:         {len(transformers):>6,}")
print(f"  AC line segments:     {len(df_lines):>6,}")
print(f"  Loads:                {len(df_loads):>6,}")
print(
    f"  Solar PV:             {len(pv_units):>6,}  ({sum(p['maxP_mw'] or 0 for p in pv_units):.0f} MW)"
)
print(
    f"  Batteries:            {len(batteries):>6,}  ({sum(b['maxP_mw'] or 0 for b in batteries):.0f} MW)"
)
print(
    f"  Synchronous machines: {len(sync_machines):>6,}  ({sum(s['maxExportP_mw'] or 0 for s in sync_machines):.0f} MW)"
)
print(f"  Wind units:           {len(wind_units):>6,}")

Summary¶

This tutorial demonstrated:

Loading a CIM XML file and exploring its structure
Extracting substations, voltage levels, transformers, lines, loads, and generation
Building a pandapower network from CIM transformer impedance data
Running a power flow analysis on the resulting model

Next Steps¶

Extend the network by adding ACLineSegment data as pandapower lines
Use Terminal and ConnectivityNode elements to resolve full bus topology
Combine CIM structure with LTDS Table 3a demand data for realistic loading
See 09-pandapower-import.ipynb for the tabular LTDS approach

Parsing CIM XML Network Data¶

What is CIM?¶

How to Get CIM Data¶

Prerequisites¶

1. Load and Explore the CIM XML¶

2. Extract Key Network Components¶

3. Parse Transformers¶

4. Parse AC Line Segments¶

5. Parse Loads (EnergyConsumers)¶

6. Parse Generation and Storage¶

7. Build a pandapower Network from CIM Data¶

8. Run Power Flow¶

9. Network Statistics Summary¶

Summary¶

Next Steps¶

Resources¶