Pelton turbine System Erlach: Part load and efficiency

System Erlach at part load: six nozzles to switch off, even on a horizontal shaft

For a Pelton turbine with several nozzles, the control system can combine the number of active nozzles with the needle position. This can keep the active nozzles at more favourable operating points when the total discharge is low. The switching sequence depends on the machine and its control system.

The more nozzles, the more steps. A conventional horizontal Pelton turbine has one or two nozzles for this. System Erlach has up to six, and the shaft is still horizontal.

For a conventional Pelton turbine with six nozzles the textbook by Brekke states an efficiency of 90 per cent at 10 to 12 per cent load, when only one nozzle is still running. When idling, a Pelton turbine needs only about 1 per cent of its discharge.

For the concept with six nozzles the 2012 paper states good operating conditions down to 5 per cent of the discharge. A single machine can therefore cover summer and winter.

Source: H. Brekke, Hydraulic Turbines, NTNU 2015, sections 4 and 5.1, pp. 16 and 22; J. M. Erlach, New Pelton Concept, Viennahydro 2012, Full Paper, section 3.1.

Pelton turbine and System Erlach: the difference at part load

Every Pelton turbine with several nozzles can switch off nozzles. System Erlach adds two things: the optional axial inlet provides approximately equal inflow conditions for the nozzles, and the nozzles may sit as far apart as the runner requires.

Source: J. M. Erlach, New Pelton Concept, Viennahydro 2012, Full Paper, sections 2.4, 2.5 and 3.1; H. Brekke, Hydraulic Turbines, NTNU 2015, p. 22.

System Erlach, efficiency per nozzle: almost the same for every nozzle with axial inlet

In the model test each of the six nozzles ran singly. With the conventional ring distributor (see Pelton turbine distributor) the efficiency depended on which nozzle it was. The paper gives as the known value from experience a deviation of plus or minus 0.3 to 0.6 per cent; measured at full load was a spread of 1.3 to 1.6 per cent below the best value between the best and the worst nozzle. With the axial inlet the six curves lay close together, about 0.4 per cent.

For engineers: The diagrams plot η / η_max, so the figures are per cent of the best value, not percentage points; the paper does not state the efficiency itself. Read from Fig. 5 and 6 (Josef Erlach, 2 October 2026): ring distributor 1.00 to 0.984 to 0.987, axial inlet 1.00 to about 0.996.

Formula as text: deviation in per cent of the best value = (1 − eta / eta_max) · 100. Quantities: eta turbine efficiency, no unit; eta_max best efficiency of the test series, no unit.

The optional axial inlet can provide approximately equal inflow conditions. The spacing of the nozzles on the runner and the control system also matter when selecting which nozzles are active.

Source: Viennahydro 2012, Full Paper, sections 2.4 and 2.5, Fig. 5 and 6; Josef Erlach, 3 October 2026 (reading of the figures).

System Erlach, nozzle spacing: 120 degrees are better than 60

If three of six nozzles are running, there are two options: three adjacent nozzles 60 degrees apart or every second nozzle 120 degrees apart. The model test shows the higher efficiency for 120 degrees, because the buckets have more time to empty. Horizontal Pelton turbine explains why conventional horizontal machines cannot use this spacing.

Source: Viennahydro 2012, Full Paper, section 2.5, Fig. 7.

System Erlach, horizontal versus vertical: measured on the same machine

The reference was a known vertical model turbine with six nozzles. The same machine with distributor, nozzles and runner was turned to the horizontal and given the new housing. The paper shows both measurements side by side and states a clear improvement for the horizontal machine at full load.

With a horizontal shaft the water leaves the buckets equally to the left and to the right. With a vertical shaft gravity pulls it to one side.

Source: Viennahydro 2012, Full Paper, sections 2.5 and 3.1, Fig. 8 and 9.

System Erlach, model measurement: test rig of Lucerne University of Applied Sciences and Arts, 0.25 per cent accuracy

The tests ran on the test rig of Lucerne University of Applied Sciences and Arts under Prof. Thomas Staubli. The laboratory demonstrated an overall accuracy of 0.25 per cent and a repeatability of 0.1 per cent. More under Pelton turbine research.

Source: Viennahydro 2012, Full Paper, sections 2.5 and 5.

System Erlach efficiency: the paper shows curves, no value in per cent

The 2012 paper shows curves that are related to the best value. It does not state an efficiency in per cent. This page therefore gives none for System Erlach.

Pelton turbine, where the per cent go: a loss balance from the literature

How the losses of a Pelton turbine are distributed was broken down by Jindřich Veselý for a high specific speed machine at its best point, by CFD simulation and model test. The figures apply to this one machine. The items are the same for every Pelton turbine, their size depends on the number of nozzles, the speed and the operating point: losses in the buckets, in the nozzle, by friction of the water on the bucket and by the air the runner drags along.

Veselý lists the item as the loss at the nozzle outlet. Josef Erlach applies it to the whole path from the inlet of the distributor to the jet: about 0.5 per cent in the nozzle at nominal opening and about 1.5 per cent in the distributor, 1.8 to 2 per cent in total. The unequal losses of the conventional distributor are the reason for the axial inlet of System Erlach. Windage grows with the cube of the speed and is spread over more power when several nozzles run.

Source: J. Veselý, Energy losses evaluation for high specific speed Pelton turbine, ČKD Blansko Engineering, pp. 255 and 257, Fig. 10 and section 4; Josef Erlach, 11 and 12 October 2026 (distributor, split).