Jun-2021

Acid gas fired reheater control

Operating as close as possible to the stoichiometric air-to-fuel ratio is advised for acid gas fired reheaters.

CLAY JONES and HARNOOR KAUR Optimized Gas Treating
ELMO NASATO Nasato Consulting

Article Summary

Modified Claus based sulphur recovery units (SRUs) require successive cooling and reheating of the process gas stream as it passes through several catalytic converter stages. Between each converter, the gas is cooled to condense and remove elemental sulphur, then reheated to allow production of additional elemental sulphur in the next stage. Figure 1 shows a typical three-converter configuration.

There are several common methods to reheat the stream including indirect steam heat, electric heaters, hot oil, gas/gas, and direct-fired reheaters.  This article focuses on direct-fired reheaters which use some of the SRU’s acid gas feed as fuel. Acid gas fired reheaters (AGFR) are burners positioned between the sulphur condenser and the next converter bed. The hot combustion gases from the burner are mixed with the main process stream in order to heat it to the desired converter temperature.

Since these reheaters are burners, they require a strategy to control the flow rate of air and fuel (acid gas). One part of this control strategy is determined by the amount of heat release needed to achieve the desired temperature rise in the process stream: this temperature sets the total amount of H2S combustion needed in the reheater.

After the temperature requirement is set, there is still a degree of freedom left in the control philosophy: should we feed the stoichiometrically required amount of acid gas such that it is all burned, should we feed an excess amount of acid gas so that the combustion products contain a 2:1 ratio of H2S:SO2, or does the best answer lie somewhere between these two approaches? Here we use a rate-based simulation to explore the process implications of this choice. Specifically, the question to be answered is what effect does the reheater’s air-to-acid-gas ratio have on overall sulphur recovery and COS generation?

Process description
The chemistry pertaining to AGFRs is generally similar to chemistry in the SRU thermal reactor (TR), although the process objectives are not identical. One of the primary objectives for the TR is to create the stoichiometric amount of SO2 that will allow the overall conversion of H2S to elemental sulphur to proceed as far as possible through the Claus reaction (Equation 1). This objective causes the optimal H2S:SO2 ratio in the TR to be close to 2:1 to match Claus reaction stoichiometry. In contrast, the primary process objective in a fired reheater is simply to liberate enough heat of combustion with a stable flame to achieve the required temperature increase in the process stream.

                [1]

As per the TR flame control strategy, in order to maintain reliable AGFR operation, it is imperative to have a proper air control system that maintains flame stability to satisfy the required temperature control setpoint. The control scheme should be programmed to allow for independent feed flow measurement on all feed streams to the AGFR burner; this includes amine acid gas and, where applicable, all fuel gas streams with steam moderation cascaded to fuel gas flow. Each feed stream will have an air demand multiplier that can be adjusted based on composition in order to provide the total flow target dependent on the cascade temperature control setpoint. Air demand requirement for each stream is then fed to a summation block to allow for ratio air control.  In gas plants, it is common to have lean amine acid gas with less than 50% H2S. For lean amine acid gas and/or turndown operation (refinery applications included), it may be necessary to co-fire the burners with supplemental fuel gas to sustain a stable flame.

Another important difference from the TR is vessel size and residence time. TRs typically accommodate most of the acid gas fed to the SRU and provide enough residence time such that slower reactions (including the Claus reaction) can approach thermodynamic equilibrium. Since fired reheaters will typically take only a small percentage of the total acid gas flow, they are designed with much shorter residence times. As a consequence of shorter residence time, kinetically controlled reactions (such as the Claus reaction) will not typically be able to achieve equilibrium in a fired reheater before the hot gases are cooled by mixing with the main process stream.

In reality there are several kinetically controlled reactions taking place in AGFRs in addition to Claus: for example, thermal splitting of H2S into H2 and S2. For the present study we will use Claus as a proxy for all of them. The extent of Claus conversion taking place in a reheater depends on the size and configuration of the equipment. If the residence time is long enough, the reaction will proceed to equilibrium. Conversely, if the residence time is very short, the Claus reaction may not occur to any appreciable extent.

Our study starts with a base case which is burning enough H2S to achieve the temperature targets for the unit operation. The amount of acid gas fed to the burner (% stoichiometric air-to-fuel ratio) is varied along with the extent that the Claus reaction is allowed to proceed. The primary design and operating decisions discussed in this article address the question as to whether the air and acid gas should be fed in stoichiometric proportions, or should the acid gas be fed in excess? How much does the design of the particular reheater vessel influence this decision?

Case study
The case study is based on a typical refinery SRU (see Figure 1). The unit feed and operating conditions are shown in Table 1. Two amine acid gas compositions were used: Test Run 1 with 92% H2S and Test Run 2 with 83% H2S. The model includes only the conversion section of the SRU. It does not include the tail gas treating unit (TGTU).

To explore the process implications of reheater operation, we ran a rate based sulphur plant simulation in SulphurPro with varying burn strategies from 30% to 90% of stoichiometric air-to-fuel ratio. The heat release requirement for a reheater does not change very much with burn strategy. Therefore, the air supply to each reheater also does not change very much; instead, we change the amount of excess acid gas sent to the reheater beyond the amount required to consume all the oxygen.

It is difficult to know the extent of Claus conversion that will occur in a reheater because it is a strong function of the size and configuration of the equipment. From a modelling perspective, the uncertainty was bounded by running the models in two different reaction modes to represent the two limiting cases. One limiting case allows all reactions to come to equilibrium by Gibbs energy minimisation which simulates a large reheater with a residence time of 0.5 seconds or more. The other limiting case prevents the Claus reaction from occurring at all, representing a small reheater with a residence time of 0.1 second or less. The behaviour of a real reheater is bounded by these two extremes.

Effect on flame temperature
The most immediate result of differing air-to-fuel ratio in the reheater is on the adiabatic flame temperature (see Figure 2). As expected, the hottest temperature is at the stoichiometric air-to-fuel ratio since deviation from this ratio implies the presence of additional unreacted gas which will act as a heat sink. The Claus reaction is endothermic at flame temperatures, so models which inhibit the Claus reaction show slightly higher flame temperatures. Note that for this case study, air-to-acid-gas ratios above 50% lead to flame temperatures that can result in refractory and burner damage.