Are you allowed to refill a percolator or hot water urn on Yom Tov?
The Federation Beis Din here in London have an incredible department called 'Fed Tech' which explores various appliances and their halachic status on Shabbos and Yomim tovim etc.
Their article on urns here addresses your question:
Can water be added on Yom Tov?
An urn which is simmerstat-controlled carries no concern with regards to adding cold water on Yom Tov as the cold water added will not affect the heating element’s operation.
When adding water to a thermostatically-controlled appliance, however, there is greater concern. This is because when a large amount of cold water is added to the urn, the heating element’s operation is changed and will come on sooner than it otherwise would have.
How do FedTech recommend that hot water be added on Yom Tov to an urn?
For urns which are thermostat-controlled, FedTech recommend adding water using one of the following four approaches:
- Add water when the device’s indicator light signals that the appliance is currently heating
- Add hot water
- Add small amounts of water so that the device will not register an immediate change. (You should try this before Yom Tov)
- Set a time switch to remove all power from the urn for a set period of time on Yom Tov. During this period, the urn should be filled. When power resumes, the urn will boil.
Simmerstat– When using a simmerstat, the lengths of the cycles are determined by the setting which the user has chosen and will repeat in a uniform manner. A dial is provided to controls these timings. Changes to the heat of the water will not affect the length of the heating cycles, as all control of the heating elements is purely time-based.
Thermostat– The length of the heating cycles will be determined by the heat of the water at any specific time. The thermostat reacts when the temperature dips below the tolerance level set by the user, and the heating element will be activated. When the desired temperature has been achieved, the heating element ceases to operate. It is unlikely that the heating cycles will be accurately predictable in their length and frequency.