All refractories need a special drying-out cycle to prevent steam explosions.  The dry out cycle is determined by the composition of the refractories.  The sensitive materials are generally the binders.  Binders are generally chemicals that react with water to form new compounds.  The commonest binder is calcium aluminate cement.  Calcium aluminate first dissolves and then combines with water to form calcium aluminate hydrates.  Different calcium aluminates can be formed at different temperatures and different conditions.  These calcium aluminate hydrates generally decompose at around 400 degrees centigrade.  The pressure of steam is relatively low at 100°C but extremely high at 400°C.  The dehydration of these hydrates is an endothermic reaction.  The initial dry out or commissioning of refractories generally takes place by a gas flame on the inside of the furnace.  The first stage of drying involves the boiling off the free water in the castable.  Whether the in the furnaces at 200° a 500° the surface of the refractory will not go above 100 degrees until all the free

water has boiled off.  The thicker the lining the greater the problem.  Drying progresses from the hot face towards the cold face.  Because of the endothermic nature of the drying process, you can have different processes active in different layers of the lining.  When the hot face has exceeded 400 degrees there can still be steam generated from deep inside the lining.  

Theoretically if you put thermocouples at different depths inside your lining you get an idea of what is going on.  In practice there are always some complications regarding the geometry of the furnace and different thicknesses in different parts of the lining.  Because of the uncertainties is it is usually a good idea to switch off the burner at intervals and listen for steam.  Steam generally makes a hissing sound.  A cold piece of glass held at the exit of a furnace, or a piece of cold steel generally shows condensation, indicating the presence of steam.  This is very useful for electrically heated processes but with gas heating there is always water in the combustion products.  

Some refractories contain clay and clay has crystal bonded water which comes off at around 600 degrees.  When firing ceramics, you also must deal with carbonates which decompose 800 degrees.

Most modern castables are called "low cement castables".  They are cheaper than conventional castables.  They have lower permeability than conventional castables.  This makes them more difficult to dry out than conventional castables.  Good manufacturers always supply dry out schedules with their materials on their data sheets.  

The risk of explosion can be greatly reduced by adding RFT fibres.  RFT stands for rapid firing technology.  These fibres are very fine organic fibres which are hollow inside.  They collapse at road low temperatures allowing steam to escape through the channels which they form.  

Obviously the drying out process is very difficult and should be done by an expert.  There are companies which specialise in this, and it is worthwhile employing them.

“Tempering” is a loose term to cover a variety of heat treatment processes.  It should not be used in ceramics, because it has a specific meaning in metallurgy.  I should rather have used the term “curing”.  The processes involved in curing/tempering are boiling off of water, dehydroxylation of hydrated minerals, phase changes of minerals, decomposition of carbonates endothermic reactions and oxidation processes.  The firing curves are slowed down at the temperature range of these reactions.  In large castings, or rather bulky castings, the reactions start on the heated surface and progress inwards as the isotherm penetrates deeper into the castings.  The heat penetration is slow during these endothermic reactions and rapid between them when only thermal conductivity determines the progress of isotherms.  After the first firing, only the phase changes of minerals are a problem, and heating can be rapid.  Where gas is released, there is a grave danger of explosion.   Dehydroxylation of calcium aluminate hydrates is the most dangerous part.  Low permeability castables increase the danger of explosion.  About 35 years ago I developed a very solid castable and made a 2,5 ton launder from it.  It exploded 4 days into its drying cycle at 400C.  The explosion was so violent that it bent the trolley made of railway line onto the floor.  It completely destroyed the drying oven.  The police and directors would not accept that a steam explosion could be that violent.  I now specify 4 day drying cycles for normal 13% porosity castables in large vessels.  Thin walled castings present no problems.  For 10mm thick castings, I fire full speed to the service temperature.  At 20mm thick I would not expect any problems except in contact with a graphite susceptor in an induction furnace.  I have only once had a customer report a steam explosion in a normal castable below 200C.  Breakage during cooling is usually encountered in pottery during the quartz inversion at 513C.  It can also occur around 200C if cristobalite is formed.  Cristobalite usually only forms at above 1100C.