Mar . 28, 2024 11:13 Back to list

Powder Activated Carbon Selection and Dosing

PAC Bid Selection Process

The process of adsorption MIB and Geosmin onto PAC takes time (> 1 hour).  Therefore, maximizing the contact time between PAC and water is critical.  PAC should be added and well mixed prior to presedimentation basins.  Chemical tracer tests should be conducted on presedimentation basins to assure there is no short circuiting within the basins, thus maximizing contact time between the PAC and water.  Based upon suspended solids analysis, hydraulic retention time (HRT) in the presedimentation and flocculation basin “count” towards PAC contact time.  PAC is removed within the first 25% of the length of the sedimentation basins.  HRT of the presedimentation basin plus flocculation basin should exceed one hour.  If HRTs are less than one hour, the PAC dose should be increased by 25%.

Not all brands of PAC remove MIB or Geosmin equally.  Bid selection criteria for PAC suppliers should be partially performance-based.  One method for bid selection includes determination of an Index Value for each PAC brand that is based upon simple laboratory PAC tests.  A PAC slurry should be prepared by adding 1000 mg of PAC to 1 liter of 0.45 mm filtered water and allowed to hydrate overnight while being mixed with a magnetic stir-bar.  Meanwhile, filter approximately 2 liters of raw water and spike with MIB and Geosmin to a representative concentration (e.g., 30 ng/L).  Fill up amber glass bottles (no headspace; 250 mL) with this water sample.  The hydrated PAC slurry will have a PAC concentration of 1 mg/ml.  Select a representative PAC dose for the performance-based experiments (e.g., 15 mg/L).  Calculate the volume of PAC slurry (VPAC) required for addition to the 250mL sample (e.g., a PAC dose of 15 mg/L would equate to 3.75 mL of PAC slurry); remove VPAC from the sample and then add VPAC of the PAC slurry to the 250mL amber bottle.  Using a magnetic stir-bar or wrist-shaker rapidly agitate the bottle containing the water sample and PAC for a desired period representative of average HRT of the presedimentation basins (e.g., 1 to 4 hours).  Immediately after the prescribed agitation period use a syringe-filter (0.2 mm) and filter the water sample / PAC mixture.  Collect the filtrate in a 100 mL amber vial (no headspace).  Conduct MIB and Geosmin analysis on a the filtrate.  Repeat for each PAC brand, and repeat for a blank (no PAC added).  Calculate the fraction of MIB remaining: C/Co where Co is the MIB or Geosmin concentration in the blank and C is the concentration after contact with the PAC.  Compute the Index Value based upon the fraction of MIB remaining (C/Co) and the unit cost of the PAC (e.g., $/lb):

Index Value = [C/Co] x [PAC unit Cost]                           Equation 1

The PAC brand with the lowest Index Value is the most cost-effective.  This assumes that there are no limitations to PAC feed rates.  For example to achieve a desired MIB removal one PAC brand may require 40 mg/L of PAC while a more expensive PAC brand may only require 30 mg/L of PAC feed.  Therefore, the actually fraction removed (C/Co) should be examined.  High PAC feed rates can increase the frequency of PAC shipments, sludge production and handling costs.

Click here for an example PAC performance-based evaluation for 8 PAC brands, provided in a blind performance test.

PAC Dosing Recommendations

After selection of a PAC brand, a dose-MIB removal nomograph should be developed.  A Dose-MIB removal plot for Norit 20B is shown in Figure 1. The graphs are developed by conducting similar experiments as for the Index value, except for just a single PAC brand and 2 or 3 different MIB initial concentrations (e.g., 30, 50, 70 ng/L).  In separate experiments the PAC dose is varied (e.g., 1, 3, 5, 8, 12, 16, 20, 25, 30, 40 mg/L).  After the prescribed contact time the samples are syringe-filtered and analyzed for MIB.  Data are plotted as C/Co versus PAC dose.  For different initial MIB concentration the C/Co versus PAC dose should overlay, since fractional MIB removal is independent of initial concentration.  All the data together should be fit with a best-fit equation (e.g., exponential fit – Equation 2).  The best-fit equation can be used directly to compute a PAC dose (see below) or used to generate a dose-MIB removal nomograph (Figure 1).

Where the raw (MIBraw) and finished (MIBfinished) water MIB concentrations are in ng/L.  For example, if the influent MIB concentration is 30 ng/L and the desired effluent MIB concentration is 10 ng/L, a 20B PAC dose of about 11.9 ng/L would be required.  Practical operating curves were generated for future use of Norit 20B by all C.O.P. WTPs (Figure 1).  The operating curves are easy to use nomigraphs what can be used in replacement of Equations 2 and 3, although the equations are more accurate than reading from the nomigraph.

Slurry storage of Norit 20B PAC at a full-scale WTP for approximately 6-months did not effect its removal efficiency for MIB in raw water.  However, ordering and storage of PAC is critical for effective MIB removal.  On-site PAC storage should be based upon maximum PAC feed rates, maximum design WTP flowrate, and deliveries every five to seven days.  A schedule of PAC deliveries should be prepared and provided to PAC suppliers at least one month in advance. PAC feed facilities should be designed for 40 to 50 mg/L of PAC.

PAC doses should be varied with flowrate and approximately weekly GC/MS analysis of raw and finished water MIB and Geosmin concentrations.  Alternatively FPA can be used more frequently to adjust PAC dosages, however, weekly confirmation by GC/MS should be included.  Costs for PAC exceed $25,000 per week during series MIB (assuming 15 mg/L, 100 MGD, $0.30/lb PAC).  Therefore conducting GC/MS analysis for MIB of raw and finished water to optimize PAC dose (Equation 3) can be extremely cost effective.  It is critical that the analytical laboratories know in advance of the MIB/Geosmin sampling and need for rapid (1 day) turnaround on the data.  All WTPs within a city should be sampled and analyzed on the same day (e.g., Monday), and PAC doses adjusted accordingly within 2-days (e.g., Wednesday).

Share

Latest news

If you are interested in our products, you can choose to leave your information here, and we will be in touch with you shortly.


Chatting

tr_TRTurkish