Chemical Pretreatment To Inhibit Scales

Feed water quality determines allowable plant operating conditions, which will dictate the optimum system recovery and overall production rate. It is necessary to make a full and detailed water analysis to identify all major anions and cations, which contribute to scale formation and general fouling. The major scaling/fouling ions are calcium, magnesium, bicarbonate, sulphate, silica, iron and barium.

Many natural waters will deposit calcium carbonate on the membrane surface if untreated. Calcium carbonate scaling potential is determined by the Langelier Saturation Index (LSI), or the Stiff & Davis Saturation Index (S&DSI) for high ionic strength waters. The risk of other sealants, such as calcium sulphate and silica are determined by measuring their ionic concentration against their known solubility products (Ksp values). Scaling results in increased pressure drop and the need for greater feed pressure to maintain constant product water output.

Dosing of chemical antiscalant reduces the risk of scaling and allows elimination of acid dosing, while maintaining plant efficiency and optimum conversion rates.

The use of an effective antiscalant will allow plant recovery to be increased to a brine LSI of up to +2.6 compared to a limitation of LSI +1.0 when using a commodity antiscalant such as SHMP or zero in an untreated system. Computer prediction programs are available from some chemical suppliers to calculate scaling potential of a range of water sources. The software allows accurate recommendations for antiscalant addition. The effect of antiscalant on scaling potential is illustrated below for a brine of LSI +2.01.The graph for the treated brine conditions indicates that the plant recovery of 70% could be increased further if plant design limitations allow. At 70% recovery 3.4ppm Antiscalant A is required to be dosed to the feedwater.

No Treatment

Treated with Phosphonate Antiscalant A

T*m<MH«tur* : | 22. j *C R*covMy<K): [ fp ' )

WductFtow: 180.0 m3/hr

CaC03CaS04B*S04SrS04 C*F2 % MAX 1CT.1 MJ OA 0« 0.0 66.0 Ul CaCOSUI 2.0t li_i

CaC03CaSOi BaS04 SrSO* C*F2 Fe SO? KKU IM 214 0.0 OJ 0.0 *A ». CaCOSLSt 2.01

Figure 2: Scaling Potential of RO brine with and without Antiscalant A In the example (Figure 2), dosing of the phosphonate based antiscalant allows safe operation of the RO plant with brine concentration of scaling species considerably in excess of normal solubility limits.

4 MINIMISING FOULING 4.1 Biofouling control

Most RO systems suffer biofouling to some degree, although this does not always severely affect performance as some biofilms remain within tolerable levels or have a high level of porosity, therefore not severely affecting the permeate flux. Biofouling potential should always be anticipated and measures taken to prevent and control biogrowth. This may require maintenance cleaning with a non-oxidising biocide or for non-potable applications intermittent biocide 'shock dosing' on-line. Laboratory analysis can be used to characterise the fouling and propose the appropriate methods of control.

Biofilm material can be scraped from fouled membrane samples for microbiological analysis consisting of basic identifications and enumeration of bacteria, fungi and yeast's. Most membrane biofilms contain both bacterial and fungal species. The physical structure of biofilms found in membrane systems can be 'gel' like or 'slimy and adhesive' with some consisting of a large ratio of polysaccharide slime to viable microorganisms. Membrane biofilms investigated in our laboratory often contain between 103 and 10s colony forming units (cfu) of bacteria per cm2 of fouled membrane.

Biocide Sensitivity Tests (BST's) have been used to evaluate the performance of selected biocides on sessile micro-organisms isolated from membrane foulant and determine optimum conditions for use. A quantitative suspension test is used to determine biocidal efficacy against bacteria and fungal species. Samples of the sessile organisms are obtained by swabbing the foulant from the membrane surface and spot or pour plate counts used to determine the efficacy of each biocide. The biocidal performance is expressed as percent kill for a known concentration and contact time. Non-oxidising biocide formulations are preferable due to the limited tolerance of polyamide to oxidising products such are chlorine or peracetic acid.

The following table outlines the multi-purpose use of a non-oxidising Biocide A as a periodically shock dosed biocide, sanitising cleaning agent or for membrane preservation.

Table 3: Membrane Biocide Treatments


Conditions of Use

Intermittent 'shock dosing' on-line

Dose 60 - 80 ppm to feed water for 4-6hrs/day

Non potable applications only

Sanitising Cleaning Agent

Recirculate 0.3% solution for 8- 10 hrs Precede and follow with alkaline surfactant

Membrane Preservative/ Biostat

Preservation period: up to 7 days: 200ppm up to 6 months: 500ppm

4.2 Maintenance cleaning

4.2 Maintenance cleaning

Maintenance cleaning will ensure optimum membrane lifetime and permeate production. Routine cleaning of membranes should always be carried out at a lower transmembrane pressure (TMP) than that used for water production. It is recommended that an operating pressure of less than 4 bar with minimal permeate flow is maintained for cleaning operations.

Cleaning practices should include periodic soaking of the membrane and the use of warm cleaning solutions up to 30°C. Membrane manufacturers' guidelines should always be followed regarding product compatibility and pH limits.

Cleaning of RO systems typically takes between 4-12 hours to perform, depending on the severity of fouling and plant size. Cleaning durations of up to 24 hours incorporating overnight soaking may be necessary if heavy biofouling is suspected. Frequency of cleaning may range from monthly cleaning cycles to an annual maintenance clean. There are many alkaline surfactants, acidic formulations and sanitising agents available in the marketplace. The complexity of the clean and number of products required for optimum cleaning conditions is wholly dependent on the composition and quantity of foulant.

5 CASE STUDY 5.1 European Paper Mill Site

Application: Production of boiler feed make-up and process water Details: The RO plant treats town mains supply to produce 500m3/d product for industrial use. Polyamide 8" brackish water membranes are installed. Prior to our investigations, pre-treatment included - sand filtration, acidification (to pH 5.5), 5 micron cartridge filters, polymer Antiscalant X and dechlorination with sodium bisulphite. The system operates at 75% recovery.

Problem: The plant was suffering from fouling and the membranes required cleaning every 2 weeks to maintain the required treated water volume.

Fouling Investigations: Membrane autopsy revealed an orange/brown foulant covering the membrane leaves and plastic spacer. Chemical composition of the major foulants was determined as 60% organics, 11.4% calcium carbonate and 15% iron oxide.

Microbiology Results: Microbiological enumerations and identifications were performed:

Bacterial Counts Fungal Counts

Membrane 3.0 xlO7 20

Plastic Spacer 8.8 x 106 2

Product Water Carrier 4.5 x 105 11

The following were identified as predominating in the foulant .

Bacteria: rod shaped bacteria, Arthrobacter Fungi: Trichoderma

Biocide tests evaluated a fast acting non-oxidising Biocide B at 200ppm and 400ppm concentration with a 30 minute contact time. A 100% bacterial and fungal kill rate was achieved at the higher concentration.

Cleaning Tests: Crossflow cleaning tests using fouled membrane samples demonstrated that an alkaline clean followed by an acidic clean would successfully remove the organics, biofilm iron and inorganic scale.

Water Analysis and Antiscalant Proposal: The feedwater supply was of good quality containing negligible quantities of iron. Iron was detected in the feedwater to the RO and the planktonic counts were 1.6 x 10® cfu/ml. Inspection of the pre-treatment plant revealed corrosion of some pipework and the inside of the sand filter vessel. A computerised scaling prediction programme calculated the brine LSI using the non-acidified feed water at 20°C, pH 7.4 and 75% recovery as +1.79. Conclusions:

• Corrosion was due to prolonged acid dosing and poor selection of materials.

• The antiscalant in use was not inhibiting scale.

• There was insufficient microbiological control at the site. Recommendations:

• It was proposed that the polymer Antiscalant X was replaced by a phosphonate Antiscalant A. This product was to be dosed at 2.78 ppm without acid adjustment.

• It was recommended that acid dosing should be ceased to eliminate the risk of further corrosion.

• The following cleaning programme was proposed: Step 1: Alkaline Surfactant A re removes organics and conditions biofilm sanitises membrane removes biofilm and other organic debris removes iron oxide and inorganic scale

Step 2: Biocide B

Step 3: Alkaline Surfactant A

Step 4: Weak Acid Cleaner D

Outcome: All of our recommendations were followed by the site. The plant is now operating well with no indication of severe biofouling, scale or corrosion on the cartridge filters or membranes. Cleaning frequency has been reduced to every 4 months.

0 0

Post a comment