Scientists have developed a new...
Tackling boron removal in seawater desalination
With growth in demand for electronic-grade silicon, and, notably, the growth of silicon solar cells, the semiconductor industry is challenged to find efficient ways to create silicon molten that is free from impurities. Pure silicon has high resistance to electrical conductivity, meaning that the silicon contains few mobile electrons. However, any small impurity, such as boron or phosphorus, present at the level of even one particle to a billion can change the electrical properties of the silicon and increase its conductivity. This phenomenon is the basis of most semi-conductor components, such as diodes, transistors, microprocessors and silicon photovoltaic cells. Deliberately creating a boron impurity inside a silicon wafer in order to change its electrical properties is known as boron doping. The process is similar to phosphorus doping. These are the most critical processes in the semiconductor industry, and the precondition is receipt of a pure molten silicon.
Different methods are implemented to create silicon free from impurities, such as removal of boron from molten silicon using CaO–SiO2-based slags, or using a steam-added plasma melting method. Regardless of the method selected, all the processes require high-purity water free of boron. Recently, governments have made strides in protecting their natural water resources for the favor of agriculture and water preservation while the industries have to solve their water shortage. This is occurring in Taiwan.
Until recently, the Formosa Petrochemical Corp. (FPCC) enjoyed freshwater from natural sources. However, the government of Taiwan reduced water rights for Formosa from natural sources to support agriculture, requiring the company to seek an independent water supply. This approach also is expanding to other countries, as high-quality desalination is a solution that industries must embrace. This conflict of supplying water for agriculture and supplying water in high purity to industry can be resolved. IDE was selected to design and supply a sea water reverse osmosis (RO) desalination plant that produces 105,000 cu meters per day of high-quality permeate, focusing on water clean of boron.
Boron removal from freshwater is a well-known process and commonly is done with boron-adsorptive resin inside an ion exchanger vessel. However, removing boron from the permeate of a seawater desalination plant to an undetectable level of 0.01 mg/L requires a different approach, and becomes a more crucial challenge when discussing a large seawater desalination plant. Although the polyamide membrane can reject boron, this process requires careful control of pH to improve the boron rejection while, on the other hand, overdosing of sodium hydroxide (NaOH) could increase the tendency for magnesium hydroxide scaling on the membrane. The design of the plant also should consider low energy consumption and minimum chemical consumption, but finding the optimal configuration and membranes becomes a challenging task.
One seawater desalination process takes the boron removal issue to the edge, using four desalination stages and a downstream boron ion exchanger that removes the boron to an undetectable level of less than 0.01 mg/L of boron.
The quantity of boron in seawater changes throughout the year and changes from one location to the other. It can range from 3 mg/L to 5.5 mg/L depending on location. The polyamide membrane can reject boron from the permeate water but it does so inefficiently. Because the boron in the water behaves similar to the dissociation of the carbonic acid, in low pH the boron is in the form of boric acid H3BO3 and the rejection of the boric acid by the membrane is limited. Therefore, in 8.2-pH sea water, a lower level of rejection of boron in the first pass is expected. Due to the presence of calcium, magnesium, and bi carbonate, the pH of the water cannot be controlled because calcium carbonate and other forms of magnesium and calcium may precipitate on the membrane.
The second RO membrane pass opens the way for a more sophisticated treatment of boron removal by increasing the pH to 10.5. The new form of boron becomes dominate in the water; it is the boric ion H2BO-3. Increasing the pH level means that more hydroxide ions (OH-) were added to the water and the electrical neutrality of the water is unbalanced. In order to balance the water, more hydrogen ions (H+) are required, and these ions are taken from the boric acid and thus form the boric ion. The second pass membrane then can easily reject more boron, as the membrane rejects better monovalent ions and the effect of the electrical field of the membrane, which better rejects negative ions due to the repulse forces. Smart management of different streams of water from the membrane can provide permeate water with less than 0.2 mg/L of boron. The design is challenged not only by the boron rejection, but also by the feed temperature. The plant is designed to work at a maximum feed temperature of 40oC, the rejection of all the ions in this temperature decreases to compensate on the low rejection due to the additional temperature stage of RO is implemented.
The semiconductor industry requires a higher level of boron rejection.
As described earlier, the rejection of boron is limited and can reach to 0.2 mg/L of boron in the permeate. This level is sufficient for agriculture but not for the semiconductor industry. Therefore, to reach the level of 0.01 mg/L of boron, a robust battery of selective boron ion exchangers removal was designed downstream the RO permeate. The selective ion exchangers were designed to take the entire desalinated water and work in counter flow mode while the chemical regeneration is done in the opposite direction. The selective ion exchanger is absorbing the boron opposite from the membrane.
The best absorption of boron happens on the acidic form of the boron. The acidic form of boron exists at a pH of 6 to 8, where all the boron is in the form of boric acid H3BO3 . Twice a year, it will be necessary to preform backwash to the resin pack bed outside the ion exchanger vessel. The purpose of the backwash is to clean the pack bed from all the fractions of resin.
Designing a robust and accurate control system is necessary to implement the proposed process. The entire plant will be fully automatic and controlled by a distributed control system. Special focus is given to the pH control, individual to each skid that is allowed to monitor the pH online with an algorithm implemented in all IDE plants. Removing boron to an undetectable level is not only a process challenge, but also a control and monitoring challenge due to the accuracy of the measurement and the unique low level of boron, which needed to be controlled.