Nov-2020

Start-up and shutdown issues in sulphur processing

Knowing the limitations of the sulphur unit and what may occur during start-ups and shutdowns helps to prevent damage.

MATHEW D BAILEY, G SIMON A WEILAND and NATHAN A HATCHER
Optimized Gas Treating Inc

Article Summary

Part 1: The SRU at turndown. The transient nature of start-ups and shutdowns arguably causes the most damage to an SRU through thermal cycling of the process equipment. Start-ups and shutdowns can be more devastating to equipment than years of steady, normal operation, and these are the very conditions often overlooked or given little thought. Thermal cycling affects the reliability of the waste heat boiler (WHB), most notably by degrading the tube sheet system, which includes the refractory, ferrules, the tube sheet itself, the fragile tube-to-tube sheet joints, and the tubes.From an initial set of parameters such as feed flow rates, composition, temperatures, and pressure, a sulphur recovery unit (SRU) is designed to meet a specific set of targets. During design, attention is given to different operating scenarios such as varying feed quality, feed rate (turndown), equipment fouling, and catalyst aging to help ensure the design is robust. Amine treating units and sulphur plants usually work well when equipment is operated at design conditions; however, equipment and instrumentation behave differently under turndown conditions, and not always in desirable ways.

It is normal for an SRU to operate below design flow rates and for operating conditions to change after construction and commissioning, and every so often during operation of the unit. Ensuring adequate performance under off-design conditions is crucial to successful operation. Through proper design, operating practices, and maintenance procedures, the reaction furnace and WHB system can have a life expectancy in excess of 20 years. However, with an inadequate design, poor operating practices, frequent cycling, or poor maintenance, it can be as short as two or three years.1 Being able to model accurately the effect of variations in feed quality, feed rate, exchanger fouling, and catalyst ageing can provide invaluable understanding of the effects of these parameters, and reveal operating conditions that will expose various SRU components to premature failure.

Case studies: turndown and operating an SRU during shutdown and start-up
A series of case studies was performed to analyse the effects of turndown on a Claus SRU. The unit studied is a typical two-stage SRU in a refinery processing both sour water acid gas (SWAG) and amine acid gas (AAG) at a combined total acid gas (CAG) design flow rate of 125 long tons per day (lt/d)(see Figure 1 and Table 1). The heat exchange units, such as the WHB and condensers, were simulated in rating mode to assess accurately the effects of operating at off-design rates. All cases were simulated using SulphurPro, a kinetic rate and heat transfer rate based sulphur recovery simulator, with the AAG to SWAG ratio fixed at 5.6 to 1.

Two cases are considered: Case A is a unit simulation in two turndown scenarios: one is at 75lt/d (60% of design) of CAG, the other at 40 lt/d (30% of design). The performance of the unit, and specifically the exchangers, is assessed at each turndown step.

Case B assesses the unit at a point during the start-up/shutdown procedure halfway between 30% turned down and hot standby (natural gas only). To achieve this condition, half the CAG at 30% load is replaced with natural gas. In other words, the overall hydraulic load is kept at 30% of design but the feed is 50:50 natural gas and CAG with CAG now at 15% of the design rate.

Results
Table 2 shows that as the unit is turned down from 100% (base case) to 60%, and then to 30% of design rates (Case A), the most noticeable changes are in the WHB operating conditions. At 30% turndown, the peak heat flux is reduced to nearly half of the base case. The cause is the severely reduced mass flux through the unit. Reduced heat flux and WHB tube wall temperatures are beneficial in reducing corrosion. However, if the rate of change from design to turndown is fast, mechanical stress from the rapid temperature changes can create differential thermal expansion and thermal shock. Without simulation, these potential hazards can be neither identified nor engineered around.

The overall Claus sulphur conversion, sulphur recovery, and CS2 in the tail gas do not seem to be highly affected by turndown. Although not directly calculated in the version of SulphurPro used in this study, pressure drop decreases rapidly with turndown.

Hydrogen in the Claus tail gas drops quite markedly from the base case design as turndown progresses. This can have a significant effect on the performance and reliability of a reduction-quench-amine type tail gas treating unit (TGTU) downstream. Per unit volume of feed gas, the reducing gas demand increases with turndown, so either more external hydrogen or more natural gas must be combusted sub-stoichiometrically in the TGTU reducing gas generator (RGG). Insufficient hydrogen increases the risk of SO2 breakthrough during turndown operations. Additionally, if the TGTU hydrogenation catalyst is not fully active, COS and CO conversion will tend to fall off first.²

Increased turndown results in considerably higher COS in the Claus tail gas. Even though rates are reduced so that more residence time is available in the hydrogenation reactor catalyst, if the TGTU catalyst is sick or there is flow maldistribution, then unconverted COS will slip through the TGTU amine system to the incinerator. If there is not a TGTU downstream of the Claus unit, stack emissions increase in direct proportion to the unconverted sulphur. Regardless of the presence of a downstream TGTU, incineration systems that are legally permitted on the basis of SO2 concentration in the stack will see an increase in SO2 at turndown; this should be considered at the design stage.

In addition to these points, there are two further complications with turndown operations. The first is the formation of sulphur fog in the sulphur condensers. The second concerns heat loss. In the case of sulphur fog, conversion to elemental sulphur is not directly affected; rather, the recovery of sulphur in the condensers suffers. At low mass velocities (< 1 lb/s.ft2), fine droplets of elemental sulphur mist evade capture by conventional mist elimination equipment, leading to reduced sulphur recover efficiency.^3,4 The risk of reaching the sulphur dew point in a downstream sulphur converter also increases, and this is compounded by increased heat loss. Besides concern for reaching the sulphur dew point in the catalyst beds, heat loss reduces the temperature in the reaction furnace and is also exacerbated at turndown. Blais et al. provided a methodology to estimate heat losses in the reaction furnace.⁵