[BUG] inconsistent sensor output behavior on disconnected branches vs appliances

[mgovers]

Describe the bug

When running state estimation with a disabled power/current sensor on a disabled branch, the output is empty (energized = 0, p_residual = 0, q_residual = 0).

However, when running state estimation with a disabled power/current sensor on a disabled shunt or load/gen, the output is not empty (energized = 1, p_residual = 6924.0000000000009, q_residual = -69280).

This behavior is unexpected

Input Data Validity

• I checked the validity of input data using the helper function (examples here).

Data Dump

Input data

{
  "version": "1.0",
  "type": "input",
  "is_batch": false,
  "attributes": {},
  "data": {
    "node": [
      {
        "id": 1,
        "u_rated": 10e3
      }
    ],
    "line": [
      {
        "id": 5,
        "from_node": 1,
        "to_node": 1,
        "from_status": 0,
        "to_status": 0,
        "r1": 1000000.0,
        "x1": 1000000.0,
        "c1": 2.0e-5,
        "tan1": 0.1,
        "r0": 1000000.0,
        "x0": 1000000.0,
        "c0": 1.0e-6,
        "tan0": 0.2
      }
    ],
    "shunt": [
      {
        "id": 7,
        "node": 1,
        "status": 0,
        "g1": 6.283185307179587e-4,
        "b1": 6.283185307179587e-3,
        "g0": 6.283185307179587e-5,
        "b0": 3.141592653589793e-4
      }
    ],
    "source": [
      {
        "id": 10,
        "node": 1,
        "status": 1,
        "u_ref": 1.05,
        "sk": 1e12
      }
    ],
    "sym_voltage_sensor": [
      {
        "id": 13,
        "measured_object": 1,
        "u_measured": 6062.1931942119563,
        "u_sigma": 0.1,
        "u_angle_measured": -1.4832085023115637e-06,
        "u_angle_sigma": 0.1
      }
    ],
    "asym_power_sensor": [
      {
        "id": 16,
        "measured_object": 5,
        "measured_terminal_type": 0,
        "power_sigma": 5.0,
        "p_measured": [1154.0, 1153.0, 1155.0],
        "q_measured": [-11540.0, -11550.0, -11550.0]
      },
      {
        "id": 17,
        "measured_object": 7,
        "measured_terminal_type": 3,
        "power_sigma": 10.0,
        "p_measured": [2308.0, 2306.0, 2310.0],
        "q_measured": [-23080.0, -23100.0, -23100.0]
      }
    ]
  }
}

Update data

{
  "version": "1.0",
  "type": "update",
  "is_batch": true,
  "attributes": {},
  "data": [
      {
          "line": [
              {
                  "id": 5,
                  "from_status": 0,
                  "to_status": 0
              }
          ],
          "shunt": [
              {
                  "id": 7,
                  "status": 1
              }
          ],
          "asym_power_sensor": [
              {
                  "id": 16,
                  "power_sigma": "inf"
              }
          ]
      },
      {
          "line": [
              {
                  "id": 5,
                  "from_status": 1,
                  "to_status": 1
              }
          ],
          "shunt": [
              {
                  "id": 7,
                  "status": 0
              }
          ],
          "asym_power_sensor": [
              {
                  "id": 17,
                  "power_sigma": "inf"
              }
          ]
      }
  ]
}

To reproduce

Run symmetric state estimation with newton_raphson

Symmetric output data (current)

{
  "version": "1.0",
  "type": "sym_output",
  "is_batch": true,
  "attributes": {},
  "data": [
    {
      "asym_power_sensor": [
        {"id": 16, "energized": 0, "p_residual": 0, "q_residual": 0},
        {"id": 17, "energized": 1, "p_residual": -14400.013604963937, "q_residual": 143960.13604963932}
      ]
    },
    {
      "asym_power_sensor": [
        {"id": 16, "energized": 1, "p_residual": -7200.0068024819693, "q_residual": 71980.068024819673},
        {"id": 17, "energized": 1, "p_residual": 6924.0000000000009, "q_residual": -69280}
      ]
    }
  ]
}

To Reproduce

Run state estimation with e.g. newton_raphson calculation method and symmetric = True

Expected behavior

The sensor output on the disabled shunt is empty

Symmetric output data (expected)

{
  "version": "1.0",
  "type": "sym_output",
  "is_batch": true,
  "attributes": {},
  "data": [
    {
      "asym_power_sensor": [
        {"id": 16, "energized": 0, "p_residual": 0, "q_residual": 0},
        {"id": 17, "energized": 1, "p_residual": -14400.013604963937, "q_residual": 143960.13604963932}
      ]
    },
    {
      "asym_power_sensor": [
        {"id": 16, "energized": 1, "p_residual": -7200.0068024819693, "q_residual": 71980.068024819673},
        {"id": 17, "energized": 0, "p_residual": 0, "q_residual": 0}
      ]
    }
  ]
}

Screenshots

No response

Additional context

This will need to be fixed in the output.hpp.

The underlying reason is that, topologically, branches are treated differently from appliances for caching reasons:

• disabling branches (and source appliances) changes the topology and therefore, the entire topology needs to be rebuilt
• disabling appliances (except sources) does not change the topology - only the Ybus parameters (shunts) or even only the actual values (loads). The topology is therefore cached. In particular, appliances will always be added to the topology and the actual output for those components is created based on the internal status - not the topological configuration.

However, the current sensor output only looks at the topological status of the measured object. This results in different behavior depending on whether the topological status is updated or not when the measured object status changes.

Instead of checking the measured object topology, the measured object status should be used to determine whether the sensor should get null_output.

Community Meeting

• I wish to discuss this bug in the bi-monthly PGM community meeting. Please join the PGM community meeting.

[nitbharambe]

Another reason to add a status attribute to the sensors.

Additionally,

• Clarify the definition of residual in documentation to include energized condition.

[nightcityblade]

Hi, I'd like to work on this. I'll submit a PR shortly.

[nightcityblade]

I need to step back from this issue. A correct fix affects both symmetric and asymmetric output fixtures and would exceed the small, focused scope I can responsibly complete in this run. The issue remains available—sorry for the churn.

[appleweiping]

Hi, I’d like to work on this issue now that #1518 has been merged, if it is still available.

I reviewed the current main, the issue discussion, and the unresolved review feedback on the closed PR #1468. My understanding of the focused scope is:

• in the power-sensor output path, use the measured component’s actual status() for source, shunt, load, and generator;
• preserve the existing topology-group behavior for branch and branch3 measurements, so a branch disconnected at only one terminal remains energized when it still contributes to the system;
• leave the current-sensor path unchanged;
• add regression coverage for symmetric and asymmetric state estimation, including enabled/disabled appliances, fully disconnected branches, and partially disconnected branch/branch3 cases;
• keep the implementation and tests focused, without unrelated refactoring.

The issue discussion also mentions documenting that residual output depends on the energized condition. Should that documentation clarification be included in this PR, or kept as a separate follow-up?

I will use OpenAI Codex interactively for repository navigation, analysis, test design, implementation, and verification. I will personally review and understand every change, remain responsible for the contribution, use signed/DCO-compliant commits, and disclose the assistance in the PR.

Does this scope match the intended fix, and may I take the issue?