WSP06151 - Laserfiche WebLink

Home Browse Search

<, NPDES1Petmit No. CO-0042161 SalinityiControl Study Twentymile Coal Compa,ny I- ..... e....) o '" ,0 ~, 3.2 <l>ption 2 - Reverse Osmosis Reduction of Salt Discharge This op~ion involves treatment of most of the pumped water to redu\le, salinitybefpre discharg~git. Three' water treatment methods, Reverse Osmosis (RO), Electrodialysis Reversal i (EDR), and Multistage Flash (MSF) distillation were initially considered for this ' treatmen~ option. RO was selected as the favored option because the technology is well develope(! and more widely applied, and the ,other technologies do not overcome the lIll\ior difficulti~ with RO, which are brine disposal, high operation and maintenance, liability lInd high capi~ cost. ' The proI!osed RO system would include chemical pretreatment, filtration, and membrane separatiob. The system design flow is I cubic foot per second (cfs), and would treat wliter , with influent TDS of 4,000 mglL to an effluent standard of 300 mglL., This system rejects 25 % of tile influent as a concentrated brine, Chemicallpretreatment includes addition of calcium hypochlorite to oxidize metals such as 4:on and manganeSe and to prevent formation of biofilm in the RO modules., Sodi\!m metabisuljiite is added !o remove residual chlorine and inhibit the formation of calcium sulfate which would foul the Ito membranes. Both reagents would be added in-line as neat solutions. Following the chemical addition steps, particulates. precipitates" and colloids would, be removed/n a two stage filtration process. First, the majority of the filterable solids will be removed In a. bank of sand filters. Sand filters would be operated under, pressure to ",inimi7.e filtration time. Effluent from the sand fJ1ters would pass through a bank of cartridge filtlirs with 5 fun disposable filter elements. After the 'filtration steps, water would enter a bank of RO membrane modules. Approximately eighty seven (87) 8" dia. x 40" long modules in stainless steel housings would be required to achieve ti\e desired effluent TDSconcentration of 300 mgll. With RO treatment about 75%pf the inflUElnt would be treated and remmed to Fish Creek, and 25,% of the influent would remain as1brine waste which will be injected in a deep well as discussed below. Disposal (If the s.a1t water brines in evaporation ponds on site is not possible because of the lack of a:suitable location for ponds. Based on a net evaporation of 12 inches per year,.a 181 acre dvaporation pond would be required for the estimated 112 gpm of brine generated by the RO ttleatment option. Even if a suitable location could be found, the evaporation pond would evdntually begin to fill with salts and a new evaporation pond location would need to be found. ' The only practical option for disposal of treatment waste waters would be deep well injection. The shallqweSt formation which would appear to be suitable for deep well injection of brines is the Dako~ Sandstone at a depth greater than 6,000 feet below surface. It is possible that isolated stlnd units could be found at shallower depths within the Mancos shale, but the chance of intercepting a sand to accept the large quantity of brine is quite low. de ,-{ .' 'l :] -~ " _,_, .,l, ^;___:u_,;,~;i- ; ;lli ~.\' -'j.- -'4 .'-, /j' i iJ AHA F"'tleaamc; 14SAI.1::N1.DOC Page 15 2:52 PM 09129(95 ;~"" ,'-_ ..;'0,-