Effluent from the SCR was first pre-filtered through a series of 5-m and 0.3-m cartridge filters (PX05-30 and PX00330, respectively; GE Water and Process Technologies, USA) and then pH adjusted to approximately 7 using HCl. Antiscalants (3.0mg/L Flocon 260) was added prior to the SRO membrane desalting stage. The SRO consisted of a 1:1 array of polyamide RO membrane modules (2540 TFC-ULP, Koch Membrane Systems, San Diego, CA) with each vessel housing three RO elements. The SRO was operated at an average feed pressure of 1217kPa (176psig) to maintain an average permeate flow rate of 5.3–5.8L/min (1.4–1.5gpm). At this permeate flow rate, the SRO feed flow rates were kept at 11.6L/min (3.1gpm) and at 7.7–8.1L/min (2.0–2.1gpm) to attain 50% and 68% SRO water recovery, respectively. The SRO unit could only be operated at a maximum of 68% water recovery in order to keep concentration polarization in the tail element within the manufacturerrecommendedlevels(i.e.,maintainsufficientcross-flow velocity of the concentrate stream). Therefore, the achievable operational upper bound of the overall pilot-scale processes was only 95% (i.e., with the PRO operating at 83% recovery).
SRO process data
were normalized per ASTM Standard D 4516-85 [16] for
permeate flow rate, salt passage, and differential pressure (for the retentate
stream across the entire collection of RO modules) with respect to the initial
process conditions (i.e., at the beginning of pilot testing). The data
normalization procedure allowed for evaluation of process variations over time
(with respect to the initial set of process conditions) after accounting for
variability of operating conditions (flows, concentrations, pressures, and
temperature), thus isolating process abnormalities (i.e., fouling, membrane
damage, etc.). The specific permeate flux (m/s-kPa) is the volumetric permeate
flux (m3/s-m2) per unit of average net driving pressure
(NDP, kPa). The NDP was calculated as the difference between the transmembrane
pressure and osmotic pressure differences between the bulk retentate and
permeate streams (i.e.,). The salt passage is the ratio of permeate salt
concentration to the log mean retentate salt concentration.
All water quality constituents were analyzed according to Standard Methods for the Examination of Water and Wastewater [17] except for the trace metals (Al, Fe, Mn, Ba, and Sr), which were analyzed according to EPA Method 200.8 [18] using inductively coupled plasma mass spectrometry (ELAN 6000 ICP-MS, Perkin-Elmer Life and Analytical Sciences, Boston, MA). Chlorine residuals in the PRO influent were measured using method 4500-Cl G [17]. Mineral saturation indices were calculated using a thermodynamic solubility model [19], using inputs derived from the water quality sampling. Membrane surface analyses included scanning electron microscopy (SEM) and energy-dispersive spectroscopy (EDS). SEM analyses were conducted using a field-emission gun (XL30-FEG, Philips Advanced Metrology Systems, Natick, MA) with EDS (EDAX, Mahwah, NJ) capabilities.
Initial SCR performance testing (Phase I pilot testing) was conducted over a period of three months to determine the impact oftheSCReffluentpHontheremovalofCa2+,Ba2+,Sr2+,Mg2+, SiO2, and dissolved organic carbon (DOC) from the primary RO concentrate, all of which were potential precursors of membrane fouling/mineral scaling. The PRO concentrate was continuously generated from CRW desalting step operated at 83% recovery (Table 1). During Phase I, the PRO concentrate was chemically demineralized in the SCR without SRO desalting (i.e., the SRO system was not in operation). The NaOH dose to the SCR feed was systematically varied over the test period to achieve SCR effluent pH within the range of 9.2–11.3. Typically, a period of about2–3hwasallowedtoreachasteady-stateSCReffluentpH.
Уважаемый посетитель!
Чтобы распечатать файл, скачайте его (в формате Word).
Ссылка на скачивание - внизу страницы.