Decarbonisation Technology August 2022 issue

molten salt from freezing on solar energy plants, storing power with heat and replacing fired heating applications. This article considers the mechanical, process, and electrical design of an EPH to successfully advance the technology into higher power considerations. Mechanical design considerations An AMSE (American Society of Mechanical Engineers) code stamp is a formal requirement for all EPHs with a volume greater than 1.5 ft 3 . In practice, specifications also require EPHs with a lower volume to be code stamped. ASME Section VIII, Div. 1 code typically requires a heater flange to be three times thicker than when using ANSI (American National Standards Institute) flanges, which have shown to be sufficient with no flange deformation reported in over 60 years. Indeed, ASME Section VIII, Div. 2 calculations show that ANSI ratings are sufficient while the Div 1 calculations are too conservative. For high gas temperature applications, the vessel shell can reach higher temperatures than the design temperature of the whole vessel due to heat radiation from the heater bundle. This has led to higher-rated ANSI flanges, even though the flanges do not reach this temperature. Ideally, the ASME should be dual rated with the shell designed for a higher temperature than the design temperature for the rest of the vessel. The first EPH specification required welding the wet side of the heater flange, which has no benefit and is higher cost. Welding the dry side is effective in all applications. Once the bundle has been welded, the welds must be hydro- tested. If one of the elements in the centre of the bundle leaks, all of the outer elements must be removed to fix the inner element weld. With a dry side weld, any element can be replaced without removing other elements. Some specifications require seamless tubing or tubing thicker than 0.035” (0.98mm). Unless the process is highly corrosive, such as for heating tail gas or ionised water, these specifications add unnecessary expense. Even though the use of copper, steel, and stainless steel is less expensive, standardising on the Incoloy 800 series balances durability with cost effectiveness and availability. Higher

alloys such as Inconel and Hastelloy are recommended for highly corrosive applications. For EPHs of 480 volt or higher, it is essential to keep the compressed magnesium oxide (MgO) free from moisture. MgO is a hygroscopic substance with high heat transfer and high dielectric strength when dry, whereas wet MgO loses its dielectric strength. Hermetic seals on the end of the heater element are required to keep the MgO inside the heater elements free from moisture. Today’s hermetic seals are done using proprietary methods for each manufacturer. However, unless it is specified, it may not be included. For hot applications, a stand-off electrical terminal housing is required to keep the wires and seals inside cool. The elements can be welded inside the terminal housing on the dry or wet side of the flange. There are advantages and disadvantages to both methods. Every aspect of electric process heating has fine details that have been resolved over time, such as Double clamping the over-temperature sensor to the heater element. This facilitates field replacement if needed. This article is meant to discuss broader EPH topics. Process design considerations Process design has improved over the years with the advent of computational fluid dynamics (CFD) giving a step improvement in computer modelling. CFD opened a black box in our understanding of EPHs: • For instance, it was thought that the hottest spot was two-thirds down the bundle, so this is where the over-temperature sensor was placed. CFD simulation showed the biggest factor is the placement of the outlet nozzle as there is a reduction in fluid velocity about two- thirds down the bundle on the same side of the outlet nozzle • In the 1990s, we found that the calculated bundle temperature of 1600°F (870°C) for heating gas to 1200°F (650°C) did not agree with the field measured temperature of 1350°F (730°C). Upon investigation, we found this was due to radiant heat from the bundle to the vessel shell and since then have incorporated radiant heat calculations for high-temperature gas applications • Another useful CFD simulation looked at the