Pilot Studies Examine RO Concentrate Management Strategies
Since the mid-1980s, the Phoenix, AZ, metropolitan area has been accumulating salts at a rate of about 1.1 million tons annually.
By Charlie He and Brandy Kelso
Since the mid-1980s, the Phoenix, AZ, metropolitan area has been accumulating salts at a rate of about 1.1 million tons annually. Local and imported sources and society's use of water are increasing the salinity of groundwater, surface water, and the reclaimed water in central Arizona. As the underflow naturally migrates toward the southwest through the area, total dissolved solids (TDS) in the groundwater increases from less than 500 mg/L in the east valley to above 2,500 mg/L leaving the valley.
Being supplemented by the brackish groundwater, especially during a drought, the surface water supplies in south Phoenix, like the supplies to the City of Phoenix's future Western Canal Water Treatment Plant (WTP), contain TDS higher than 800 mg/L for at least half of the time in the last 10 to 15 years. Municipal and agricultural use of the local water resources subjects these waters to anthropologic-induced increases in the TDS concentration. The resulting high-TDS reclaimed water percolates to the water table or is discharged to local surface waters that are hydrogeologically connected to the groundwater.
Given the magnitude of the salinity issue associated with almost all available water sources, water providers in central Arizona cannot solve the issue individually and have to work together to assess the problem and develop regional strategies for managing it. To accomplish this, the Central Arizona Salinity Study (CASS) was initiated in 2001. CASS began through a cooperative partnership between the U.S. Bureau of Reclamation (USBR), the City of Phoenix, and the Sub-Regional Operating Group (SROG) represented by the cities of Glendale, Mesa, Phoenix, Scottsdale, and Tempe, Arizona. Collaborating with partners such as SROG, USBR, Water Research Foundation, and WateReuse Research Foundation, the City of Phoenix continued its efforts approaching salinity management solutions in multiple steps.
This article presents a series of desalination and concentrate management studies initiated by the city for its future Western Canal WTP and other master planned facilities. It depicts the trajectory of an environmentally accountable utility's progress in maximizing the water recovery while lowering the chemical, energy and environmental impacts.
The technology proposed for the Western Canal WTP salinity control is low pressure reverse osmosis (RO). Most of the two-stage or three-stage brackish RO systems often operate at approximately 75 to 90 percent recovery, generating around 10 to 25 percent concentrate. Due to the difficulty in concentrate disposal, the need for high-recovery RO systems and low-energy concentrate management strategies are highly desirable.
Defined as the product water relative to the input water flow, the recovery of brackish water RO is often limited by sparingly soluble constituents, such as silica, barium, strontium and calcium. To maximize RO recovery, the limiting scale-forming constituent must be identified, removed and/or controlled. Two strategies were identified in the City of Phoenix Water Quality Master Plan with respect to the scaling control for the future Western Canal WTP:
- Maximize the scaling control using optimized scale inhibitor dosing strategy, i.e., high doses of innovative scale inhibitor products with optimized pH control;
- Remove the scale-forming constituents through a chemical precipitation step to allow higher recovery through a subsequent secondary RO system.
Based on literature and preliminary water sampling and testing results, it was anticipated that the first strategy has the potential to improve the overall recovery from 85 percent to up to 90 percent. The second strategy is promising to enhance the recovery up to 96 percent. The city embarked on a series of pilot testing to further investigate each of the strategies.
Phoenix Area Membrane Pilot Study (2003-2006): Maximize Recovery Using Scale Inhibitors
In 2003, USBR and the City of Phoenix awarded a project to Carollo Engineers to investigate the treatability of the brackish surface and groundwater sources for the future Western Canal WTP. The main objective of the study was to maximize the RO recovery using a newly developed dendrimer scale inhibitor.
Several types of scale inhibitor chemicals exist, including threshold inhibitors, distorting agents, and dispersants. Threshold inhibitors retard precipitation of salts by binding with ionic charges. Distorting agents alter crystal growth to make them weaker and more prone to fracture, which in turn makes the scale that does form easier to clean from the membrane surface. Dispersants add charge to the crystals, causing them to repel one another and making it more difficult for scale buildup to form.
Dendrimers, one type of dispersant, use very long repeating chains to capture inorganic substances. The physical characteristics of dendrimers appear to have corrected many of the issues associated with polyacrylate and phosphonate types of scale inhibitor. For example, high TDS, iron and manganese can cause interference with conventional scale inhibitors. This issue is not suspected for dendrimer-based antiscalants due to their unique geometry and potential of increased solubility, resulting in higher application rates, reduced scale formation, and improved production efficiency.
The pilot plant was operated for approximately 160 days on surface water (between September 2004 and February 2005) and approximately 130 days on groundwater (between July 2005 and January 2006). During the surface water testing, an ultra-filtration (UF) pretreatment was used. The testing concluded that operating at 85 percent recovery was sustainable, but 90 percent recovery was not feasible. The decrease in membrane performance as indicated by a loss of permeate production and salt rejection was observed after less than 30 days of operation at 90 percent recovery.
During the groundwater testing, pretreatment with UF was not used. The RO plant performed much better initially, but also showed scaling/fouling in the RO membranes after less than 30 days when operating at 90 percent recovery and a dendrimer scale inhibitor dose of 3.2 mg/L. Periodic membrane cleaning was not able to sustain permeate production. A membrane autopsy showed the presence of scale containing silica, iron, calcium and aluminum in the tail end membrane element.
Western Canal WTP Water Quality Sampling and Testing Study (2007 - 2009): Achieve Sustainable High Recovery
A couple years later, upon the completion of the Western Canal WTP Master Plan, the city initiated a pilot testing study at the same Western Canal site to further investigate the feasibility of this potentially effective concentrate volume reduction strategy. The main objective of the completed pilot study was to demonstrate the performance of a conventional lime softening Intermediate Concentrate Chemical Stabilization (ICCS) process in removing the limiting sparingly soluble constituents (i.e., silica, calcium, etc.). This technology uses a solids contact clarifier to precipitate calcium carbonate and magnesium hydroxide, and co-precipitate the sparingly soluble ions that can cause membrane scaling and limit membrane recovery. It stabilizes concentrate to allow further recovery in a subsequent secondary RO system.
During the Phoenix pilot study, brackish groundwater and brackish "surface" water (actually a brackish groundwater and surface water blend) were desalted using the same RO system. Overall recovery of 94 to 95.5 percent was achieved on both sources. The ICCS reactor performed consistently well. An average of 70 to 80 percent total silica was removed by ICCS through both the groundwater and blend water phases. The reactor exceeded the treatment goal of 50 percent silica removal established by membrane modeling. The ICCS reactor successfully responded to the water source change when switching between groundwater and surface water.
Effective as it is in removing the membrane scale-forming constituents, the conventional softening-based ICCS reactor requires a significant amount of lime (1.2 to 1.5 grams per liter) to raise the concentrate pH to above 10.5. The process also generates a large volume of residuals, which in turn has to be handled using gravity thickener and/or dewatering centrifuges.
Western Canal Tailored Collaboration (2009 - 2010): Lowering the Energy and Chemical Consumption
Being fully committed to its stewardship of the Southwest region's water resources, the City of Phoenix didn't stop its investigation on minimizing the concentrate volume. Through a tailored collaboration program with the Water Research Foundation, the city extended the Western Canal conventional lime ICCS testing in an effort to lower the total water costs and environmental impacts associated with concentrate minimization.
Literature review concludes that a modified pellet softening reactor could reduce chemical addition, footprint, and residual volume. In a pelletized softening reactor, the concentrate is pumped in an upward direction, maintaining the pellet bed in a fluidized state. Precipitation occurs on the surface of microsand particles, which are suspended in the reactor. In order to crystallize the target component on the pellet bed, a driving force is created by a reagent dosage and pH adjustment. By selecting the appropriate process conditions, co-crystallization of impurities can be achieved. The pellets grow and move toward the reactor bottom. At regular intervals, a quantity of the largest fluidized pellets is discharged from the reactor and fresh seed material is added. After atmospheric drying, readily handled and virtually water-free pellets are obtained.
The Western Canal Tailored Collaboration testing focused on demonstrating the performance of the pelletized softening ICCS while lowering the chemical usage and energy consumption. Preliminary results from the ongoing pilot study demonstrated that the pelletized softening technology achieved over 50 percent removal of silica on average. The secondary RO performed stably well at 50 percent recovery on groundwater primary RO concentrate, achieving an overall system recovery of 94 percent.
In summary, the conventional softening ICCS reactor requires a significant amount of chemical. The process generates a large volume of residuals, which in turn must be handled using thickeners and/or centrifuges. An ideal application for the conventional ICCS process may be at a brackish surface water treatment facility that already has residuals handling systems in place.
The pelletized ICCS technology could be ideal for wellhead or centralized groundwater desalting facilities. It reduces the chemical usage and greatly reduces the residuals volume. Most of the residuals, in the form of pellets, are non-hazardous and easy to handle. The pellets have a promising potential for beneficial reuse as a concrete aggregate, whitening agent, soil conditioning agent, and material supplement for roof shingles, etc. If generated from a reclaimed water application, thus containing nutrients such as phosphorus, the pellets may be used as a slow-release fertilizer.
To fully evaluate the environmental impacts of the conventional and pelletized softening ICCS technologies, the Western Canal Tailored Collaboration study included a Life Cycle Assessment (LCA) based on conceptualized 10 mgd brackish groundwater desalination facilities using primary RO, conventional or pellet softening ICCS, secondary RO, and evaporation ponds in the area of Phoenix.
The LCA is essentially a "cradle-to-grave" assessment used to evaluate the full environmental implications of products in every life cycle stage (i.e., materials extraction and production, manufacturing, transportation, use, and ultimate fate of the product).
As illustrated in the figures below, LCA concluded that the pelletized softening ICCS provides a feasible option that could reduce the environmental impacts associated with the concentrate management. When comparing just the ICCS process instead of the entire facility, such reductions are even more prominent.
About the Authors: Charlie He is a senior project engineer with Carollo Engineers and the Southwest Regional Lead of Carollo's Research and Development Group. Brandy Kelso, PE, is Civil Engineer III Team Leader with the City of Phoenix Water Services Department.