Proceedings 36th New Zealand Geothermal Workshop 24 - 26 November 2014 Auckland, New Zealand UNUSUAL ARSENIC SULFIDE SCALING IN A BINARY TEST PLANT Kevin Brown 1,2 ,Jenny Webster-Brown 1,2 , Maja Rasmussen 2 and Simon Addison 3 1 GEOKEM, PO Box 30-125, Barrington, Christchurch, 8422, NZ 2 University of Canterbury, Private Bag 4800, Christchurch 8140, NZ 3 Mighty River Power, PO Box 245, Rotorua 3040, NZ kevin@geokem.co.nz Keywords:Arsenic sulfide, scaling, binary plants, geochemical thermodynamics. ABSTRACT During the planning period for the Ngatamariki geothermal power station, a pilot plant was constructed by Mighty River Power to investigate options for the design of the final production plant. During the binary plant testing at lower temperatures, a deposit appeared at 85ºC in the test section after the retention tank. This deposit was not present during a similar test at 90ºC. The deposit was orange-brown in colour, very porous and had deposited a significant thickness (~0.15mm) during the 18 days of the test. SEM investigation of the deposit showed that it was composed of mainly arsenic and sulphur with iron and antimony as minor constituents. Geochemical thermodynamic modelling indicated that the fluid was significantly undersaturated with respect to amorphous orpiment (As 2 S 3 ), at 85 o C, but was possibly saturated with respect to arsenopyrite (FeAsS). SEM-EDS analysis of the deposit indicated that the atomic ratio of As:S was approximately 1:1, more akin to the ratio in arsenopyrite than the 1:1.5 that would be expected for As 2 S 3 , but the Fe:As ratio was only 0.11:1. This could be the first recorded occurrence of an (Fe-depleted) arsenopyrite depositing in surface geothermal plant. Due to the formation of this scale, the reinjection temperature at Ngatamariki was designed to be always at or above 90ºC. 1.0 INTRODUCTION During the planning period for the Ngatamariki geothermal power station, a pilot plant was constructed by Mighty River Power to investigate options for the design of the final production plant. The pilot plant was designed to be able to test a variety of dual flash and binary options for development. The test plant also included a retention system to simulate fluid residence time in reinjection pipelines and reinjection wells. Removable pipe sections throughout the test plant were included in the pilot plant to investigate scaling and corrosion for the conditions tested. 2.0 PILOT PLANT Only the binary section of the pilot plant is described here. A full description of the pilot plant is given in Addison and Brown (2012). A diagram of the binary section of the test plant is shown in Figure 1. Separated brine and steam were supplied to the pilot plant from an adjacent well (NM7) and separator. The steam was passed through a heat exchanger and exsolved gases were vented to the atmosphere. The condensed steam, less the vented gases, was added to the brine and then passed through a further heat exchanger. The steam heat exchanger is normally only designed to remove the latent heat of the steam with a very small decrease in temperature of the condensate. The condensate normally has a low pH due to the residual dissolved gases and when added to the brine lowers the pH of the brine. This assists in reducing the likelihood of silica polymerisation as the pH is reduced before the silica saturation index is exceeded. However this reduction in pH increases the likelihood of antimony and arsenic sulphides (Brown, 2011) depositing within the preheater, reducing efficiency of the unit as deposition occurs. The combined brine and condensate was then fed to a hold up vessel which provided a ~55 minute retention time to simulate the time spent in reinjection pipelines and reinjection wells. Removable pipe sections labelled as “test sections” in Figure 1, were located at strategic positions to enable scaling and corrosion to be investigated for different operating conditions. After each different test condition, the test sections were removed, and the internal surfaces investigated by optical microscopy and SEM, The SEM analysis also included EDX analysis of the elemental composition of any scale deposited. Figure 1: Diagram of the binary section of the pilot plant 2.1 Conditions tested Three trials were conducted on the binary test rig. Steam and brine from NM7 were separated at 13 barg (195ºC). Steam flow of 0.4 t/hr was condensed in the heat exchanger and gases were exsolved at a pressure of 12.5 barg. The condensate exit temperature was 180 – 185 ºC. The condensate was then mixed with a brine flow of 1.6 t/hr and the combined brine and condensate (2.0 t/hr) had a temperature of 185 – 190 ºC. The exit temperature of the mixed heat exchanger was 95ºC, 85ºC and 90 ºC in the first, second and third tests respectively. 2.2 Results of the pilot testing The first test (exit temperature = 95ºC) ran for 11 days and the other two tests ran for 18days each. The pH 25 of the brine and condensate mixture was measured as 5.6 in the brine exiting the mixed heat exchanger. In the three tests, the deposits observed in the test sections 5 and 6 were identical in all three tests, as the conditions at these positions were the same.