Chile’s smallest registered operating solar plant: Concón#

PMGD PFV Concón (CEN ID 2082) is the minimum positive potencia_maxima among the 757 operating solar entries in the 2026-10-31 CEN snapshot: 0.1 MW. This is not a claim about every small private, net-billing or off-grid system in Chile. The installer reports 118 kWp DC; that is not the registry power field.

The same experimental counter used for Icor is attempted on real imagery. It returns no resolved module count, not zero modules. Glare and unresolved frame lines prevent a defensible inventory; no capacity-to-module conversion or manually supplied row count is used to fill the gap.

Download both offline Chile examples and audit exports. See the evidence and reproduction guide.

import json
from dataclasses import replace
from pathlib import Path

import matplotlib.pyplot as plt

from energy_analysis.module_counting.chile import audit_site_image, load_chile_study
from energy_methods.pv_module_counting import count_regular_pv_modules

HERE = Path(globals().get("__file__", Path.cwd() / "example.py")).resolve().parent
DATA = HERE / "data/chile_modules"
manifest, ranking = load_chile_study(DATA)
audit = audit_site_image(DATA, "concon")
print(ranking[["id", "nombre", "potencia_maxima"]].head(6).to_string(index=False))
print(
    json.dumps(
        {k: audit[k] for k in ("status", "image_candidate_count", "whole_plant_image_count")},
        indent=2,
    )
)
  id                           nombre  potencia_maxima
2082                  PMGD PFV CONCON            0.100
 308      PMGD PFV TECHOS DE ALTAMIRA            0.150
1786                   PMGD PFV CAIMI            0.200
2177    PMGD PFV JARDIN SOLAR PETORCA            0.210
2430              PMGD PFV EL CATURRO            0.210
1727 PMGD PFV UTFSM VALPARAISO VALDES            0.215
{
  "status": "unresolved_grid",
  "image_candidate_count": null,
  "whole_plant_image_count": null
}

1. Review the inspection scope and actual optical evidence#

The rectangle encloses the visible array-like structures; it is manually selected and is not an operator-validated boundary. The registry point lies elsewhere in the industrial site. The source metadata probe reports a 2026-07-02 acquisition, 0.34 m source resolution and 0.3 m service sampling. These metadata describe one probe, not guaranteed uniform coverage of the image. No sharpening or synthetic detail is introduced.

fig, axes = plt.subplots(1, 3, figsize=(16, 6), layout="constrained")
for ax in axes[:2]:
    ax.imshow(audit["rgb"], interpolation="nearest")
    ax.contour(audit["mask"], levels=[0.5], colors=["#00d8eb"], linewidths=1)
    ax.set_axis_off()
axes[0].set_title("Concón · inspection scope in cyan")
axes[1].set(xlim=(170, 410), ylim=(400, 115), title="Native-pixel zoom · glare obscures frames")
smallest = ranking.head(6).iloc[::-1]
axes[2].barh(
    smallest.nombre.str.replace("PMGD PFV ", "", regex=False),
    smallest.potencia_maxima * 1000,
    color="#1d8794",
)
axes[2].set(
    xlabel="CEN maximum-power field (kW)", title="Six smallest positive operating solar entries"
)
axes[2].tick_params(axis="y", labelsize=8)
fig.suptitle("Concón: 100 kW registry entry · automatic module count unresolved", fontsize=17)
fig.text(
    0.02,
    0.015,
    "© 2026 Esri and its licensors · Esri, Vantor, Earthstar Geographics, "
    "GIS User Community. All rights reserved.",
    fontsize=8,
)
fig.set_layout_engine("constrained", rect=(0, 0.06, 1, 0.94))
Concón: 100 kW registry entry · automatic module count unresolved, Concón · inspection scope in cyan, Native-pixel zoom · glare obscures frames, Six smallest positive operating solar entries

2. Check that minor parameter changes do not manufacture a count#

This small sensitivity check is not an accuracy assessment. Each unresolved run remains None. A future non-null result would still need visual review and held-out labelled modules before being treated as a reliable count.

result = audit["result"]
attempts = []
for blue in (0.005, 0.01, 0.025):
    trial = count_regular_pv_modules(
        audit["rgb"], audit["mask"], config=replace(result.config, blue_excess=blue)
    )
    attempts.append({"blue_excess": blue, "image_candidates": trial.estimated_modules})
print(json.dumps(attempts, indent=2))
[
  {
    "blue_excess": 0.005,
    "image_candidates": null
  },
  {
    "blue_excess": 0.01,
    "image_candidates": null
  },
  {
    "blue_excess": 0.025,
    "image_candidates": null
  }
]

3. Preserve the evidence gap#

No independent module inventory has been retrieved for this plant. Obtain an as-built module schedule or a glare-free orthophoto/UAV survey. The installer photographs confirm the PV technology but do not show every module unobscured; they are linked in the guide, not substituted for a full-site measurement.

plt.show()

Total running time of the script: (0 minutes 0.660 seconds)

Gallery generated by Sphinx-Gallery