In short: Guatemala’s regulator publishes the limits as hard numbers and the results as colour maps. The limit is worth knowing: under NTSD Article 56, a low-voltage user may lose at most 12 hours over six months urban, 14 rural, across 6 or 8 interruptions — with interruptions under three minutes and force majeure excluded entirely (Art. 54). Cross the line and Article 58 compensation is automatic, whether or not the distributor was at fault. The one per-distributor quality figure CNEE prints numerically is voltage regulation in 2024: EEGSA 100.00% of readings in tolerance, DEOCSA 79.65%, DEORSA 82.22%. If you are choosing where to work remotely, that 20-point gap is the honest signal — and it is about voltage, not blackouts.

This page is about reliability standards and your entitlements. For who to call when the lights go out and where the outage maps are, use the power outages page; for what electricity costs, the electricity page.

Why this page exists

Every relocation guide tells you Guatemala has power cuts. None of them tells you the number at which a power cut stops being weather and starts being a breach with a price attached. That number is published, in a 114-page technical norm almost nobody outside the industry opens, and it is genuinely actionable: it tells you what to expect, what to log, and what to claim.

It also has a hard limit, which this page states up front. CNEE publishes the measured outage results only as municipal choropleth maps with no numeric labels. The 2024 Compendio Estadístico presents “Duración Promedio en Horas, de Interrupciones Durante 2024” and “Frecuencia Promedio en Cantidad de Veces” as coloured maps. No per-distributor TIU or FIU figure can be read from them, so none is published here. What follows is the regulatory ceiling, the one numeric per-distributor comparison that exists, and the entitlements — not a league table of measured outage hours, because the regulator does not publish one in a citable form.

The ceiling: how much outage is legally acceptable

The Normas Técnicas del Servicio de Distribución measure continuity over continuous six-month control periods (Art. 53), using four indices (Art. 55):

IndexWhat it measuresWhose experience
FMIK“la cantidad de veces que el kVA promedio de distribución sufrió una interrupción de servicio”Network-wide (global)
TTIK“el tiempo total, en horas, en que cada kVA promedio estuvo fuera de servicio”Network-wide (global)
FIUnumber of interruptions suffered by each individual userYou (individual)
TIUhours of interruption suffered by each individual userYou (individual)

FIU and TIU are the ones that describe your house. Both are per six-month period.

Article 56 — the individual tolerances (from month 13 of the Régimen stage, all users)

Your connectionFIU urbanFIU ruralTIU urban (hours)TIU rural (hours)
Baja Tensión (a normal home)681214
Media Tensión46810
Alta Tensión36

Source: NTSD Art. 56, read 2026-09-17. The Alta Tensión row is published as a single FIU/TIU pair without an urban/rural split.

Annualised, that is roughly 24 hours a year urban and 28 rural for a normal home — doubling the semester figure. That doubling is arithmetic performed here, not a CNEE-published annual figure.

Article 56 — the global tolerances (Régimen stage, low-voltage users)

Cause of interruptionFMIK urbanFMIK ruralTTIK urban (h)TTIK rural (h)
Attributable to distribution2.53.5810
Attributable to causes external to distribution4 (urban & rural)12 (urban & rural)

The external-causes row is published as a single pair of values spanning both columns, not as separate urban and rural figures.

The earlier Transition-stage values were looser — FMIK 3 / 4 and TTIK 10 / 15 for distribution-attributable interruptions, 5 and 20 for external causes.

What does not count, at all

This is where most people’s intuition and the statistics part company. NTSD Article 54:

“Se considerará como interrupción toda falta de servicio de energía eléctrica en el punto de entrega. Para efectos de estas Normas, no se considerarán las interrupciones menores de tres minutos; así como las que sean calificadas como casos de fuerza mayor.”

And a 2006 amendment (Resolución CNEE 18-2006) adds a third exclusion:

“Las interrupciones causadas por Fallas de Larga Duración no serán incluidas en la evaluación de los índices individuales, Frecuencia de Interrupción por Usuario y Tiempo de Interrupción por Usuario.”

So three categories vanish from the indices:

EventCounted?Practical effect
The 90-second blip that reboots your router and drops your callNo — under three minutesThe outage that most damages remote work is statistically invisible
A tropical storm taking the region downNo — force majeureThe worst days are excluded by design
A long-duration failureNot in FIU/TIUExcluded from the individual indices since 2006
A four-hour unplanned cut on a clear TuesdayYesThis is what the ceiling is built for

The consequence for anyone planning video calls: the sub-three-minute blip is the outage class that ruins remote work and the one class the regulator does not count. Size a UPS for it; do not expect the statistics to warn you about it. Nothing in the NTSD helps you here, which is exactly why a battery backup on the router — and, at Lake Atitlán or in the Petén, a Starlink dish with its own power resilience — is the real mitigation.

What you are owed when the ceiling is crossed

This is the part worth reading twice. NTSD Article 58:

“Se aplicarán indemnizaciones a los Usuarios en caso de superarse cualquiera de las tolerancias establecidas por causas atribuibles o no al Distribuidor, a partir de la Etapa de Régimen.”

Compensation is owed whether or not the distributor was at fault. That is unusual and it is the sentence to quote if you are ever told a cut was “not our responsibility”.

How it is calculated, from the norm’s own formulas:

TermDefinition as published
INII“Indemnización para ser distribuida individualmente, (Q)”
INIG“Indemnización para ser distribuida globalmente, (Q). Cada Usuario recibe una Indemnización proporcional a su consumo semestral de energía eléctrica”
ENS UsuarioD Usuario × [ (TIU – TIU límite) / 8760 ] — a quantity attached to you, scaled by how far over the limit you went. The norm’s own definition of D Usuario was not captured for this page, so it is left untranslated rather than guessed
CENS“El Costo de Energía No Suministrada es diez veces el valor del Cargo Unitario por energía de la Tarifa simple para Usuarios conectados en baja tensión sin cargo por demanda (BTS), de la ciudad de Guatemala, del primer día y primer mes del período de control evaluado”

Two design details that matter in practice:

  • The unserved energy is priced at ten times the retail tariff. That multiplier is what gives the mechanism teeth.
  • You get one or the other, not both. Article 58: “A los Usuarios que se les aplica una Indemnización individual, no les corresponderá una Indemnización global.”

Note also Article 57’s staging: individual FIU/TIU control applies to one hundred per cent of users from the thirteenth month of the Régimen stage, so a low-voltage household is inside the individual regime, not only the network-average one.

The one per-distributor comparison CNEE publishes as numbers

In the CNEE’s 2024 Compendio Estadístico, Gráfica 7 reports the results of the voltage-regulation measurements the distributors themselves submitted:

DistributorReadings inside toleranceReadings outside tolerance
EEGSA100.00%0.00%
DEORSA (Energuate)82.22%17.78%
DEOCSA (Energuate)79.65%20.35%

Source: CNEE, Compendio Estadístico, 2024 data, published August 2025, Gráfica 7, p.75, “Fuente: cálculos con base en Información regulatoria remitida por agentes.” Read 2026-09-17.

On territory: CNEE’s chart names the distributors, not their service areas. Our own electricity page describes EEGSA as serving Guatemala City, Sacatepéquez and parts of Escuintla, with Energuate covering the rest — that description comes from the site, not from this CNEE chart, and the same page carries distributor user counts that CNEE’s 2024 figures contradict. Treat the service-area split as approximate.

Read this precisely. It is a Producto Técnico result — the quality of the electricity delivered — not a Servicio Técnico result. It says roughly one measurement in five on the Energuate networks fell outside the permitted deviation from nominal voltage, while EEGSA’s fell outside none. It does not say Energuate customers lose more hours of supply; CNEE does not publish that numerically.

What “outside tolerance” means is set by Article 24. From month thirteen of the Régimen stage, permitted deviation from nominal voltage is:

Voltage levelUrbanRural
Baja Tensión±8%±10%
Media Tensión±6%±7%
Alta Tensión±5% (urban & rural)

As with the interruption tolerances, Alta Tensión is published as one figure, not an urban/rural pair.

And: “Se considera que la energía eléctrica es de mala calidad cuando, en un lapso de tiempo mayor al cinco por ciento del correspondiente al total del Período de Medición, las mediciones muestran que la Regulación de Tensión ha excedido el rango de tolerancias establecidas.”

Why a remote worker should care about voltage and not only outages: sustained low voltage is what kills refrigerators, inverters and laptop power bricks, and it does so without ever producing an outage you would notice or log. If you are buying appliances for a house on the Energuate network, a voltage regulator is a more defensible purchase than it looks from the outside — and the 20-point gap in the table above is the published reason.

The commercial standards you can hold them to

These are the deadlines to cite in a complaint. All from the NTSD, Régimen stage.

StandardLimitArticle
Percentage of users filing complaints (R)≤ 5% per semesterArt. 64 a)
Average time to process a complaint (TPPR)≤ 10 daysArt. 64 b)
Metering error, users up to 11 kW3% maximumArt. 64 c)
Metering error, users above 11 kW2% maximumArt. 64 c)
Notice before a planned interruption48 hoursArt. 64 d)
Reconnection after you pay24 hours maximumArt. 69 c)
New connection, no network modification30 days maximumArt. 69 a)
New connection requiring network modification4 months maximumArt. 69 b)
Billing-error complaintsResolved on the next bill (or the one after, if filed within 15 days of issue)Art. 69 d)

Article 69 d) also carries a rule worth knowing if a billing error keeps happening: the distributor must check whether the same error hit other users and fix those “resolverlos inmediata y automáticamente a todos los afectados, sin esperar nuevos reclamos”, and the same error recurring within six months for the same users counts as reincidencia.

To use any of this, the route is the one already documented on the power outages page: complain to the distributor first with dates and times, then escalate to CNEE if the answer is unsatisfactory.

Who actually serves you, and how big they are

You do not choose your distributor — your address does.

DistributorApproximate users, 2024Largest single department
EEGSA~1.58 millionGuatemala, ~1.2 million users = 76% of EEGSA’s total
DEOCSA (Energuate, west)~1.45 millionHuehuetenango, ~237,000
DEORSA (Energuate, east)~0.95 millionAlta Verapaz, ~149,000

Source: CNEE Compendio Estadístico, p.41, 2024 data. Nationally there are 19 distributors, and these three serve roughly 94% of regulated users. Social-tariff users are 90% of EEGSA’s base and over 95% of both Energuate companies'.

Average monthly consumption differs enormously, which is itself a clue about the network:

DistributorGeneral (kWh/month)ResidentialCommercialIndustrial
EEGSA2051485,98631,989
DEORSA117976,36432,564
DEOCSA87755,88931,532

Source: CNEE Compendio Estadístico, Tabla 1, p.14 (2024 data).

The compendium’s own explanation for why Energuate tariffs and quality differ is density, not management. Of the distribution component of the tariff it says: “en EEGSA tiene una participación significativamente menor que en las empresas DEOCSA y DEORSA, esto se atribuye a la dispersión de los usuarios que atienden dichas distribuidoras.” Longer, thinner rural networks cost more per user to run and are harder to hold inside a voltage band. That is the structural reason behind both the tariff gap and the voltage-compliance gap.

What this means if you are choosing where to work from

  • On the EEGSA network you are on the distributor that reported zero voltage readings outside tolerance in 2024, on the cheapest tariff, and with by far the highest user density — CNEE puts about 1.2 million of EEGSA’s ~1.58 million users in the department of Guatemala alone. It is the default answer for anyone whose income depends on an uninterrupted video call.
  • On an Energuate network, budget for a voltage regulator on anything with a compressor or a sensitive power supply, and treat the published tariff gap and the published voltage-compliance gap as two symptoms of the same structural cause.
  • Everywhere, size a UPS for the sub-three-minute event, because it is the one the regulator does not count and the one that drops your call. UPS sizing for a Guatemala home office covers the arithmetic, and solar and battery options cover the larger version of the same problem.
  • Log your outages with dates and times. The Article 58 entitlement is real, it does not require proving fault, and the only thing that makes it usable is a record.

What would improve this page

A numeric, per-distributor TIU and FIU series. CNEE computes these — the compendium says so — but publishes them as municipal colour maps. If CNEE releases the underlying table, or an open-data endpoint, the reliability comparison in this page’s first section becomes a measured one rather than a regulatory ceiling. Until then, this page publishes the ceiling and says so.

Sources

All read 2026-09-17.

Spanish quotations are transcribed from the CNEE PDFs named above. Annualised figures are stated as arithmetic performed on CNEE’s semester-based limits and are labelled where they appear.

Frequently Asked Questions

How many hours of power cuts per year is normal in Guatemala?

The regulator does not publish a national average, but it does publish the ceiling. Under Article 56 of the Normas Tecnicas del Servicio de Distribucion, a low-voltage user in the mature regulatory stage may lose at most 12 hours of supply per six-month control period in an urban service area, or 14 hours rural, across at most 6 urban or 8 rural separate interruptions. Doubling the semester figures gives roughly 24 urban and 28 rural hours a year as the regulatory ceiling — that doubling is our arithmetic, not a CNEE figure. Medium-voltage users get a tighter ceiling (8 urban / 10 rural hours) and high-voltage tighter still (6 hours, 3 interruptions).

Do short outages count?

No. Article 54 of the NTSD excludes two whole categories from the reliability indices: interruptions shorter than three minutes, and anything classified as force majeure. A 2006 CNEE resolution added a third exclusion — long-duration failures are left out of the individual FIU and TIU indices. So the flicker that reboots your router at 2 pm never appears in the statistics, and neither does the storm that took the region down for two days.

Is EEGSA more reliable than Energuate?

On the one comparison CNEE publishes as numbers, yes, and by a wide margin — but it is a voltage-quality comparison, not an outage comparison. In the CNEE’s 2024 statistical compendium, 100.00% of EEGSA’s voltage-regulation measurements fell inside tolerance, against 79.65% for DEOCSA and 82.22% for DEORSA — meaning roughly one measurement in five on the Energuate networks was outside the permitted deviation from nominal voltage. CNEE publishes the actual outage figures (TIU and FIU) only as coloured municipal maps with no numbers on them, so no per-distributor outage hours can be quoted from it.

Can I get money back for power cuts in Guatemala?

Yes, and the striking part is that fault does not matter. Article 58 of the NTSD says compensation applies where any of the tolerances is exceeded ‘por causas atribuibles o no al Distribuidor’ — for causes attributable to the distributor or not. It is calculated from a quantity the norm calls D Usuario, the hours by which the limit was exceeded, and a cost of unserved energy set at ten times the unit energy charge of the simple low-voltage tariff in Guatemala City on the first day of the control period. The norm’s own definition of D Usuario was not captured for this page, so it is not translated here. Where the breach is measured individually you get an individual payment; where it is measured globally, each user gets a share proportional to their semester consumption.

How much notice must I get before a planned outage?

48 hours. Article 64, literal d) of the NTSD lists ‘Falta de Notificacion de Interrupciones Programadas: 48 horas, antes del inicio la interrupcion’ among the commercial-quality tolerances. That is the standard the distributor is measured against, and it is why the distributors publish a planned-outage schedule at all.

How fast must the distributor reconnect me after I pay?

Within 24 hours. Article 69, literal c) of the NTSD: once the cause of the cut has been resolved and the user has paid what is owed, the distributor is obliged to reconnect ‘dentro de un plazo maximo de veinticuatro horas’. The same article gives 30 days maximum for a new connection that needs no network modification, and 4 months where the network must be modified.

What counts as a voltage problem rather than an outage?

Article 24 of the NTSD sets the permitted deviation from nominal voltage at the delivery point. In the mature stage, from month thirteen, a low-voltage urban service may deviate by 8% and a rural one by 10%; medium voltage 6% urban and 7% rural; high voltage 5%. Supply is deemed to be of poor quality when measurements show the deviation exceeded those bands for more than five per cent of the total measurement period. This is what the 2024 EEGSA-versus-Energuate comparison measures — brownouts and over-voltage, not blackouts.

Which distributor will I have, and can I choose?

You cannot choose; your address decides. CNEE’s 2024 compendium counts approximately 1.58 million users on EEGSA, 1.45 million on DEOCSA and 0.95 million on DEORSA, out of 19 distributors nationally, with those three serving roughly 94% of regulated users. Guatemala department alone accounts for about 1.2 million users, 76% of EEGSA’s total.