Hydrogen-cooled turbo generator
A hydrogen-cooled turbo generator is a turbo generator with gaseous hydrogen as a coolant. Hydrogen-cooled turbo generators are designed to provide a low-drag atmosphere and cooling for single-shaft and combined-cycle applications in combination with steam turbines. Because of the high thermal conductivity and other favorable properties of hydrogen gas this is the most common type in its field today.
Based on the air-cooled turbo generator, gaseous hydrogen went into service as a coolant in the rotor and the stator in 1937 at Dayton, Ohio, in October by the Dayton Power & Light Co allowing an increase in specific utilization and a 99.0% efficiency.
The use of gaseous hydrogen as a coolant is based on its properties, namely low density, high specific heat, and the highest thermal conductivity (at 0.168 W/(m·K)) of all gases; it is 7-10 times better at cooling than air. Another advantage of hydrogen is its easy detection by hydrogen sensors. A hydrogen-cooled generator can be significantly smaller, and therefore less expensive, than an air-cooled one. For stator cooling, water can be used.
Generally, three cooling approaches are used. For generators up to 300 MW, air cooling can be used. Between 250-450 MW hydrogen cooling is employed. For the highest power generators, up to 1800 MW, hydrogen and water cooling is used; the rotor is hydrogen-cooled, the stator windings are made of hollow copper tubes cooled with water circulating through them.
The generators produce high voltage; the choice of voltage depends on the tradeoff between demands to electrical insulation and demands to handling high electric current. For generators up to 40 MVA, the voltage is 6.3 kV; large generators with power above 1000 MW generate voltages up to 27 kV; voltages between 2.3-30 kV are used depending on the size of the generator. The generated power is left to a nearby station transformer, where it is converted to the electric power transmission line voltage (typically between 115 and 1200 kV).
To control the centrifugal forces at high rotational speeds, the rotor is mounted horizontally and its diameter typically does not exceed 1.25 meter; the required large size of the coils is achieved by their length. The generators operate typically at 3000 rpm for 50 Hz and 3600 rpm for 60 Hz systems for two-pole machines, half of that for four-pole machines.
The turbogenerator contains also a smaller generator producing direct current excitation power for the rotor coil. Older generators used dynamos and slip rings for DC injection to the rotor, but the moving mechanical contacts were subject to wear. Modern generators have the excitation generator on the same shaft as the turbine and main generator; the diodes needed are located directly on the rotor. The excitation current on larger generators can reach 10 kA. The amount of excitation power ranges between 0.5-3% of the generator output power.
The rotor usually contains caps or cage made of nonmagnetic material; its role is to provide a low impedance path for eddy currents which occur when the three phases of the generator are unevenly loaded. In such cases, eddy currents are generated in the rotor, and the resulting Joule heating could in extreme cases destroy the generator.
An on-line thermal conductivity detector (TCD) analyzer is used with three measuring ranges. The first range (80-100% H2) to monitor the hydrogen purity during normal operation. The second (0-100% H2) and third (0-100% CO2) measuring ranges allow safe opening of the turbines for maintenance.
Hydrogen has very low viscosity, a favorable property for reducing drag losses in the rotor; these losses can be significant, as the rotors have large diameter and high rotational speed. Every reduction in the purity of the hydrogen coolant increases windage losses in the turbine; as air is 14 times more dense than hydrogen, each 1% of air corresponds to about 14% increase of density of the coolant and the associated increase of viscosity and drag. A purity drop from 97 to 95% in a large generator can increase windage losses by 32%; this equals to 685 kW for a 907 MW generator. The windage losses also increase heat losses of the generator and the associated cooling problems.
The absence of oxygen in the atmosphere within significantly reduces the damage of the windings insulation by eventual corona discharges; these can be problematic as the generators typically operate at high voltage, often 20 kV.
The bearings have to be leak-tight. A hermetic seal, usually a liquid seal, is employed; a turbine oil at pressure higher than the hydrogen inside is typically used. A metal, e.g. brass, ring is pressed by springs onto the generator shaft, the oil is forced under pressure between the ring and the shaft; part of the oil flows into the hydrogen side of the generator, another part to the air side. The oil entrains a small amount of air; as the oil is recirculated, some of the air is carried over into the generator. This causes a gradual air contamination buildup and requires maintaining hydrogen purity. Scavenging systems are used for this purpose; gas (mixture of entrained air and hydrogen, released from the oil) is collected in the holding tank for the sealing oil, and released into the atmosphere; the hydrogen losses have to be replenished, either from gas cylinders or from on-site hydrogen generators. Degradation of bearings leads to higher oil leaks, which increases the amount of air transferred into the generator; increased oil consumption can be detected by a flow meter associated to each bearing.
Presence of water in hydrogen has to be avoided, as it causes deterioration to hydrogen cooling properties, corrosion of the generator parts, arcing in the high voltage windings, and reduces the lifetime of the generator. A desiccant-based dryer is usually included in the gas circulation loop, typically with a moisture probe in the dryer's outlet, sometimes also in its inlet. Presence of moisture is also an indirect evidence for air leaking into the generator compartment. Another option is optimizing the hydrogen scavenging, so the dew point is kept within the generator manufacturer specifications. The water is usually introduced into the generator atmosphere as an impurity in the turbine oil; another route is via leaks in water cooling systems.
The flammability limits (4-75% of hydrogen in air at normal temperature, wider at high temperatures), its autoignition temperature at 571°C, its very low minimum ignition energy, and its tendency to form explosive mixtures with air, require provisions to be made for maintaining the hydrogen content within the generator above the upper or below the flammability limit at all times, and other hydrogen safety measures. When filled with hydrogen, overpressure has to be maintained as inlet of air into the generator could cause a dangerous explosion in confined space. The generator enclosure is purged before opening it for maintenance, and before refilling the generator with hydrogen. During shutdown, hydrogen is purged by an inert gas, then the inert gas is replaced by air; the opposite sequence is used before startup. Carbon dioxide or nitrogen can be used for this purpose, as they do not form combustible mixtures with hydrogen and are inexpensive. Gas purity sensors are used to indicate the end of the purging cycle, which shortens the startup and shutdown times and reduces consumption of the purging gas. Carbon dioxide is favored as due to very high density difference it is easily displaced by hydrogen.
Hydrogen is often produced on-site in electrolyzers, as this reduces the need for stored amount of compressed hydrogen and allows storage in lower pressure tanks, with associated safety benefits and lower costs. Some gaseous hydrogen has to be kept for refilling the generator but it can be also generated on-site.
See also 
- Development of world's largest hydrogen-cooled turbine generator
- A chronological history of electrical development from 600 B.C.
- TCD analyzer