Improving the performance of LIFAC FGD in Chinese boilers

Rapid economic development in China has brought with it a corresponding increase in air pollution from energy use. Environmental protection is therefore evolving into a critical factor in determining the rate of economic and industrial growth.

Coal accounts for 90 per cent of China's overall primary fuel resources. 75 per cent of all electricity generation is fuelled by coal. Although there is a wide variation in air pollution across the country. more than 40 per cent of China is polluted by acid rain, largely caused by SO2 discharge. Damage to crops, materials and human health is significant – for 1995 it was reckoned at 110 billion RMB yuan, equivalent to two per cent of the GNP. The Chinese government has paid increasing attention to the control of SO2 pollution in recent years. Comprehensive environmental legislation in 1997 obliged power plants using coal with a sulphur content higher than one per cent to install desulphurisation units.

So far, the capacity of imported FGD units in operation accounts for only 2.2 per cent of installed coal burning capacity. International companies have turned their attention to the large market in China and national research is being conducted to develop a Chinese FGD technology. But limited financial resources – except in a number of special economic zones – is delaying the installation of FGD units.

Selection factors

A number of different FGD technologies have to be evaluated to find the most economical and technically flexible solution for each case. The regulatory standards have to be met, but they are expected to change in the coming years, so the selected process may need to meet tighter standards than those in force at the time of selection. The availability of the required sorbent material and the production of waste water or solid waste may restrict the choice of technology. Limited space for installation of the FGD plant may increase the construction costs in some cases. Most importantly, the total costs of desulphurisation must be optimised for the operating hours of the plant, costs of consumables, the evaluation of the investment cost compared to operating cost and the preferences of the customer.

The LIFAC process

The LIFAC (limestone injection into the furnace and activation of calcium oxide) flue gas desulphurisation process (Figures 1, 2) is an improved limestone injection process where post furnace activation of unreacted lime is carried out in a separate flue gas humidification chamber. The process can remove 75 to 85 per cent of the SO2, it is relatively simple and the capital cost is typically less than 50 per cent of the cost of the wet FGD method. It uses inexpensive and abundant reagents, which reduces life span costs.The process requires an activation reactor which is installed between the boiler and the electrostatic precipitator.

In the first stage finely pulverised limestone is pneumatically blown into the boiler furnace where the temperatures are 900 to 1250°C. Chemical reactions in the furnace include breakdown or calcination of the limestone (CaCO3) and reactions of SO2 and S03 with the resulting calcium oxide (CaO). This mixture of particulate reaction products and fly ash is carried into the activation reactor by the flue gas stream.

CaCO3 = CaO + CO2

2CaO + 2SO2 + O2 = 2CaSO4

CaO + SO3 = CaSO4

In the second stage the remaining calcium oxide reacts with water to form calcium hydroxide (Ca(OH)2) in the activation reactor; this reacts rapidly with SO2 to form calcium sulphite (CaSO3), some of which oxidises to calcium sulphate (CaSO4).

CaO + H2O = Ca(OH)2

Ca(OH)2 + SO2 = CaSO3 + H2O

2CaSO3 + O2 = 2CaSO4

The product is a dry powder, most of which is disposed of with fly ash from the electrostatic precipitator, and the rest separated from the bottom of the activation reactor. Part of the fly ash from the ESP and the reactor bottom is sent back to the activation reactor. Fly ash from the process may be used for road construction, land filling, mine back filling or raw material for bricks. There is no waste water, but the process can use some of the plant's waste water in the humidification reactor.

The first step provides 25 to 35 per cent SO2 removal by injection of pulverised limestone into the furnace. The investment cost is very modest, typically about 10 per cent of the cost of the complete system. The humidification reactor is the heart of the system. The humidification of flue gas and recycling of dry ash increase the SO2 removal to 75 per cent. The investment cost of the second stage is roughly 85 per cent of the cost of the complete system. The performance of the LIFAC process can be improved by slurry ash recycling, (Figure 3) which increases SO2 removal to 85 per cent at about 5 per cent of the total investment cost. These stages can be added to the existing installation when required without replacing the original equipment, allowing some flexibility in planning the investment as well as in meeting changing emission standards. In addition, some flexibility is gained in the choice of fuel supply.

The total cost per unit of sulphur removed by the LIFAC process is low, and its simplicity keeps total capital costs low. The major part of the equipment can be manufactured locally, which lowers the cost further. The equipment is small in size, making it easy to retrofit or to incorporate into the design of new power plants where the space constraints are an important consideration.

Chinese projects

The first two projects in China were the systems for Xiaguan power plant in the city of Nanjing. The old plant was dismantled and replaced by new plant consisting of two 125 MWe units, each with a LIFAC desulphurisation system. The process is designed to remove 75 per cent (Ca/S 2.5) of SO2. The most recent project is the LIFAC FGD in Qianqing power plant (Figure 4) in Zhejiang province near Hangzhou. The power plant has a new 125 MWe coal fired unit. The required SO2 removal in this case is 65 per cent (Ca/S 2.35).

One problem was that the amount of fly ash taken from the reactor bottom was higher than specified. In the LIFAC process the by-product is collected from the ESP ash handling systems and from the bottom of the reactor. The latter ash is conveyed into the vertical flue gas duct upstream of the reactor. Finer particles of the ash are carried by the flue gas back to the reactor while the coarser ones drop into a hopper at the bottom of the vertical inlet duct. From there they are dropped through double dampers onto a lorry. In this case less by product was collected from the ESP and more fly ash was collected from the reactor bottom. This meant that less fly ash was recirculated back into the activation reactor and less free reagent was utilised, affecting the efficiency of the system. The second, more important issue, was that guaranteed SO2 removal values were being achieved only occasionally. In order to solve the these two problems the CFD modelling tool Ardemus, developed by Fortum with the Technical Research Centre of Finland (VTT), was used.

In the first case the reason turned out to be flue gas volume, which was slightly below the design figure. This caused lower flue gas velocity and the recycled fraction of by-product was not caught by the flue gas flow in the vertical inlet duct. To solve the problem the inlet duct was narrowed to optimise the quantity of by product falling out of the system and the amount of by product recirculation. The flow in the inlet duct was simulated by Ardemus and the design of the modification was based on the result.

In the second case the reason for the deteriorated SO2 removal was inefficiency of the limestone injection process. The most important factor here is the calcination of CaCO3 to CaO, and the most important parameter affecting the calcination is the furnace exit gas temperature (FEGT). Too high temperatures cause sintering of the CaO and too low temperatures cause poor conversion of CaCO3 to CaO. In this case the FEGT turned out to be lower than specified, because of non-adjusted boiler parameters. The desulphurisation efficiency is also highly dependent on the mixing of the calcinated sorbent and the flue gas which in turn depends heavily on the aerodynamics in the furnace.

Furnace modelling

To establish the furnace exit gas temperature/mixing efficiency relationship, the furnace was simulated using Ardemus. The actual boiler design, fuel specification and operating conditions were used in the simulation. The temperature profile of the Nanjing Xiaguan boiler showed low furnace exit gas temperatures due to non-optimised adjustments of the furnace in the early part of the operation. Trajectories of limestone particles showed that the particles did not reach high enough temperatures. The low temperatures affected both the calcination of limestone and the total plant electrical efficiency. The FEGT was adjusted accordingly; the parameters affecting FEGT are primary air temperature, burner tilt angle, burner levels in use, sootblowing interval and limestone injection level.

The primary air temperature affected the FEGT strongly. Lowering the primary air temperature in Nanjing Xiaguan reduced the coal/air (C/A) ratio resulting in a lower flame temperature. When this happens the combustion area moves upward in the furnace causing higher furnace exit gas temperatures. In addition steam temperature increases when FEGT increases. The FEGT was also increased by using upper burner levels in the furnace and tilting the burners upwards. Less sootblowing decreases heat transfer in the furnace, which increases the FEGT.

The limestone can be injected on two different levels. The upper injection level is intended to be used on full load and the lower on low load. As the FEGT was low, the lower level was used also on full load, in order to increase the temperature window of the limestone particles. On finding the optimum boiler parameters, SO2 removal efficiency clearly exceeded the guarantee value of 75 per cent with the guaranteed Ca/S molar ratio of 2.5. When optimising boiler parameters for improved SO2 removal, plant efficiency increased due to higher steam temperatures. Figure 5 shows the temperature field, and Figures 6 and 7 show two views of the limestone particulate trajectories of Nanjing Xiaguan boiler after optimising the boiler conditions. For tangential firing, the aerodynamics of the sorbent mixing region result in special design considerations: depending on the strength of the tangential vortex in the furnace the limestone is mixed with the flue gas in different ways. The optimum location of sorbent injection with respect to temperature and mixing conditions can be found for a specific boiler using Ardemus modelling.