Chemical Water and Wastewater Treatment III

Coagulants, flocculation time and mixing intensity were examined in laboratory studies using a synthetic fulvic acid water and three Swedish water supplies. Alum and polyaluminium chlorides (PACs) of various chemistries were effective when used under favorable pretreatment conditions. Alum and PACs had approximately the same optimum dosages for the fulvic acid water when expressed as mg Al added per mg TOC. An exception was the Ca PAC had a lower optimum dose at higher pH. The PACs were effective at flocculation times as low as 5 min while 10 min was needed with alum, especially in cold waters. Dissolved air flotation (DAF) performance improved slightly with increasing flocculation mixing intensity. DAF performed better than settling for aquatic humic or algae type supplies especially with lower flocculation times, and performed better for a river water with turbidity under cold water conditions. Excellent DAF treatment was achieved with Ca, Si, and sulfate PACs.

The Sjölunda wastewater treatment plant comprises mechanical, biological and chemical treatment. The chemical treatment step includes flocculation and flotation. The concentration of suspended solids has been higher than expected when the plant was built. In this study factorial designs were used in combination with laboratory tests in order to identify factors of significance for the removal of suspended solids. The primary aim of the project was to reduce the suspended solids to about 10 mg/l.

The investigation showed that the type of water, secondary effluent from trickling filters or an activated sludge process, and the flocculation energy were important for the process performance. The dosage of alum and polymer and the dosage point of the polymer were also significant. A high flocculation energy, i.e. a case with relatively small flocs, improved the overall removal. It also seems possible to decrease the concentration in the tertiary effluent to below 10 mg/l.

This pilot study was designed to evaluate the efficacy of membrane microfiltration in treating secondary wastewater effluent. An 8 liter per minute pilot was operated at the City of Houston’s Southwest Wastewater Treatment Plant. Membrane permeates were of similar or better quality than filtrates from the plant’s conventional sand filters. Good removal of particulate contaminants, including coliform bacteria, was observed. In comparison with the plant’s filtrate, permeate quality was also much less variable. Operating the pilot in a crossflow configuration did not improve the permeate quality and did not retard the increase in transmembrane pressure in comparison with dead-end operation. Chemical cleaning of the membrane with caustic solution was required approximately once every 5 days. Preliminary results indicate that coagulation pretreatment in conjunction with membrane microfiltration can also be used to remove phosphorus from wastewater.

The influence of different doses of aluminium sulphate and polyaluminium chlorides when cleaning a thin and a thick sewage on the residual turbidity, concentrations of TOC, ortho-P, non-ortho-P, SS, and alkalinity and sludge volume produced was studied.

Both the alkalinity consumed and the amount of ortho-phosphate precipitated decreased with increasing basicity of the coagulant. At low coagulant doses no or little alkalinity was consumed. When the coagulants were added above a threshold concentration the alkalinity consumed as well as the amount of sludge produced increased strongly with the dose used.

When the thick water was cleaned to a constant phosphorus concentration the type of coagulant giving lowest sludge volume depended on the amount of phosphorus removed. At high removal rates the lowest basicity coagulants gave the lowest sludge volume. When the residual P concentration was 1.5 mg/1 the coagulant with OH/A1 of 1.1 gave the lowest sludge volume.

When waters were cleaned to a constant turbidity the sludge volume decreased somewhat with increasing basicity of the coagulant.

This paper summarizes a series of research projects in State Key Laboratory of Environmental Aquatic Chemistry (SKLEAC).

The chemical species in polyferric solutions were examined by timed complexation spectrophotometry to classify them as Fe _a , Fe _b and Fe _c . The hydrolysity B* taken as a principal parameter is emphasized. The classification and evolution patterns of hydrolyzed Fe(III) solutions are examined.

Short term precipitation in polyferric chloride solutions can be retarded by addition of a sort of phosphate stabilizer which contributes to this flocculant both a commercial possibility and a higher efficiency.

To examine the reuse of municipal waste water, besides jar tests in the laboratory, pilot tests of advanced treatment after the biological oxidation process were carried out. The polyferric chloride flocculant was used and compared with other coagulants. A high efficiency of removal of turbidity and color was obtained.

This paper is concerned with the performance of a relatively new form of Fe(III) coagulant, Polyferric Sulphate (PFS), which contains a range of pre-formed Fe(III) hydrolysis species with both a high positive charge and medium and large molecular weights. Preliminary laboratory and pilot-plant experiments were undertaken to evaluate the performance of PFS, in comparison with conventional coagulants such as ferric sulphate (FS), for the coagulation of surface waters containing substantial concentrations of algae and/or colour, and natural organic matter.

The performance of PFS was found to be consistently superior to other coagulants in terms of the removal of turbidity, algae, colour, UV absorbance and TOC, THMFP and the residual concentration of coagulants. In addition, sludge production and headloss development in the subsequent sand filter process were both lower with PFS than FS, and the influence of low water temperatures (e.g. 4°C) on the coagulation performance with PFS was less than that with FS and aluminium sulphate (AS). The reasons for the superior performance of PFS and the effects of pre-ozonation and low water temperatures on coagulation behaviour are discussed in the paper.

Iron (VI) ion FeO ₄ ^2- possesses properties which make it potentially useful in certain areas of water purification as an oxidizing, coagulant, flocculant and bactericidal agent. Due to its high instability, ferrate (VI) has not been extensively produced and studied. In this paper, we recommend a synthesis method of alkali and alkaline-earth stable ferrates which can be easily produced industrially; these stable ferrates are characterized by the formula M (Fe, X)O_4. Applications of these compounds, in particular K₂(Fe, S)O₄, in water treatment have been investigated.

An iron (II) and hydrogen peroxide mixture, also called ‘Fenton’s reagent’, produces an extremely strong oxidizing agent the hydroxyl radical OH●. Recent publications by several authors have shown that use of Fenton’s reagent could be considered for use in water treatment, due to its very effective oxidation of specific organic substances like pesticides. Another advantage of Fenton’s reagent is that it does not produce bromates. Iron (III), which is the other reaction product of Fenton’s reagent, could certainly act as coagulant in the treated water.

There are some differences between coagulation with Fe³⁺ and that with Fenton’s reagent. Optimal dose of iron in the case of Fenton’s reagent depends on interactions between the alkalinity (pH) and H₂O₂/Fe²⁺ ratio. For the same concentration of humic substances in model humic water (5 mg/l of HS) the optimal dose varied from 40 μmol/l to 110 μmol/l of Fe, depending on the other variables. Residual colour, residual UV absorbance (254 nm) and residual iron also depend on interactions between initial alkalinity, coagulation pH, iron dose, and H₂O₂/Fe²⁺ ratio. No values of optimal pH or iron dose can be given or recommended as guide values at this time. It seems that optimal iron doses are lower when Fenton’s reagent is used rather than iron (III). The H₂O₂/Fe²⁺ ratio should not drop below 0.5 but, on the other hand, there is only a slight improvement when it exceeds 0.5.

Water is a strange substance with many physical and chemical peculiarities. As water is a very good solvent it is essential for all kinds of life but creates also problem for life. The phrase, “Use water wisely”, is obvious and worn. But why don’t we trust it? The political goal, however, is now to create a sustainable environment that creates opportunities for a variety of different species.

Soft, acidic, and coloured (humic substances) surface waters with a low turbidity are commonly used for the drinking water supply in Norway. For these waters, coagulation/direct filtration combined with a process for corrosion control is one of the most interesting water treatment alternatives.

In the experimental work described here, coagulation/direct filtration of water from Lake Jonsvatnet was investigated in a pilot plant using three different coagulants, alum (ALG), prepolymerized aluminium chloride (PAX14), and a calcium containing prepolymerized aluminium chloride (Ca-PAX). Magnafloc LT20 as a filter aid and three different combinations of chemicals for corrosion control purposes were also included in the pilot plant experiments: 1) lime/CO₂, 2) CaCO₃-slurry/CO₂/NaOH, and 3) CO₂/NaOH (using Ca-PAX).

A dual media (hydroanthracite on top of sand) gravity filter with a total depth of 0.85 m was used in the experiments. Headloss and effluent quality in terms of pH and turbidity were monitored continuously, and colour, Ca, alkalinity, and Al were analyzed from water samples taken throughout each filter run.

Optimum type of coagulant and combinations of chemicals for corrosion control were evaluated in terms of effluent quality, utilization of chemicals, cost of treatment, headloss, and length of filter run versus filter load.

Satisfactory treated water quality with respect to the Norwegian water quality standards was obtained with all the applied combinations of coagulants and chemicals for corrosion control. Lake Jonsvatnet is relatively low in humic content and moderate in corrosivity. The treated water quality was approximately: colour = 3 mgPt/1, pH = 8.0, alkalinity = 1.0 meq/1, and Ca = 20 mg Ca/1. However, the utilization of chemicals, filter performance and treatment costs were largely affected by the applied combinations of coagulants and chemicals for corrosion control.

The results from the investigation show that the most cost effective choice of coagulant, coagulant dosage, and coagulation pH is largely affected by the type of corrosion control process and can be considerably different for a process without water treatment for corrosion control. The results may be used to design and optimize a more cost effective overall treatment process for these types of waters.

Results from pilot studies of the use of ozonation and biologically active GAC media to upgrade two existing direct filtration facilities are summarized. A comparison of preozonation followed by GAC filter media to post-filtration ozonation followed by a GAC contactor is made based on various aggregate measures of organic matter including DOC, THMFP, HAAFP, AOC and odors. The results show that ozonation and biologically active GAC media improve organic matter removal. For one plant, preozonation and post-ozonation at similar ozone to DOC ratios and GAC contact times produces very similar water quality while at the other plant, the post-ozonation option is significantly better. The choice of the location for ozonation and GAC at these facilities will ultimately depend on water quality goals and disinfection needs.

Granular filter design should rely not only on process evaluation but also on overall economic considerations. A logical model is introduced for optimizing a deep bed filter and minimizing experimentation. The model consists of two stages, a process stage and an economic stage. Conventional filtering materials as well as basalt and tuff are evaluated in single, dual, and mixed media configurations. Tests were performed in a pilot plant treating Lake Kinneret water. It is concluded that the optimization process should involve a process optimization stage followed by an economic optimization stage. The filter process parameters, run length and produced water volume per run do not provide enough data for an economic evaluation of different filter beds. Single deep bed filters of basalt or tuff can remove turbidity as efficiently as single, dual, and mixed media filters containing conventional filtering materials, and produce cheaper water. Determination of optimal media parameters should be carried out at the optimal mode of backwashing and chemical treatment. Process efficiency indicators, such as volume of produced water per filter run, are poor economic indicators and should be used carefully.

The British often use the expression “value for money”. In this paper we will try to demonstrate that this expression is very valid for chemical wastewater treatment. Different aspects of wastewater treatment will be examined, such as:

The Klagshamn wastewater treatment plant is operated with extended nutrient removal. The process configuration is based on pre-precipitation and postdenitrification in a single sludge activated sludge system. As the nitrogen removal at the plant is strongly affected by the sludge age, the capacity of the system should be increased if the load of particulate material on the biological treatment is decreased.

In order to improve the suspended solids removal in the pre-precipitation process, a full-scale test with polymer addition was performed at the plant The test clearly showed that the concentrations of suspended solids in the primary effluent decreased when a polymer was added to the pre-precipitation step. The decreased load of particulate material on the biological treatment resulted in an increased sludge age and an improved nitrification capacity. Model simulations showed that this improvement allowed the flow to the biological treatment to be increased by about 15 %.

This paper presents results concerning three wastewater treatment plants combining biological treatment (by upflow biofiltration) and chemical treatment in order to achieve:

The results show advantages of such an association which makes it possible to reach high effluent quality and reliability with respect to stringent standards (C, N, P).

A process for rapid clarification of sewage has been developed, in Australia, by the CSIRO and extensively tested in pilot scale studies in Melbourne, Sydney and the UK. The results of the pilot studies were used to design a 5 Ml/day prototype of the SIROFLOC™ sewage treatment process (STP). This was constructed at the Malabar sewage treatment works (STW), in Sydney, and has been successfully operated since October 1992.

The process utilises finely divided particles of magnetite combined with an inorganic coagulant to aid the rapid separation of colloidal and suspended solids from sewage. The magnetite and inorganic coagulant are recovered and reused within the process, thus minimising the operating costs. The use of magnetite allows for the rapid coagulation and clarification of sewage; raw sewage can be treated to a high quality within 15 minutes.

Since the process is a coagulation process, it is ideally suited to the removal of insoluble matter from sewage. The 5 Ml/day prototype plant at Malabar has operated continuously to produce treated sewage containing 30 mg/l TSS and 6 mg/l oil and grease, equivalent to reductions of up to 90 %. Phosphate removal was better than 90%, consistent with the use of an inorganic coagulant and total coliforms were reduced by up to 3 logarithms. Significant reductions in heavy metals were achieved by the process, with removal efficiency dependent on the particular metal. Soluble contaminants present in the sewage, such as ammonia, were not significantly removed by the process. Biochemical and chemical oxygen demand (BOD and COD, respectively) were generally reduced by at least 50 %, consistent with the removal of their insoluble fraction from the sewage.

Öresundsverket in Helsingborg is the first municipal wastewater treatment plant in Sweden, designed and revamped for biological nitrogen removal. Different alternatives of phosphorus removal were used during the first two years of operation. The main purpose of this paper is to illustrate how the biological nitrogen removal has been influenced by the different phosphorus removal alternatives.

Pre-precipitation had a positive effect on the nitrification rate, because of a lower carbon to nitrogen ratio. The other processes tested gave equivalent nitrification rates.

Denitrification rates determined with acetate as single carbon source gave equivalent rates to all tested processes. Denitrification with methanol had an adaptation time of at least 1.5–2 months.

The characteristics of the activated sludge varied substantially, diluted sludge volume index (DSVI) between 100 and 400 ml/g. The recirculation in the biological treatment is an important factor that influenced DSVI. A higher amount of easily biodegradable carbon had a positive effect on DSVI.

Pre-precipitation gave high DSVIs with large variations from week to week.

The Point Loma Waste Water Treatment plant in San Diego operates as an advanced primary treatment plant at a mean capacity of 190 MGD (720 000 m³/day). The discharge from the plant is through a four mile long outfall tunnel into the Pacific Ocean. Many scientists have documented that the BOD discharge does no harm to the marine environment, which is why most efforts have been focused on increasing the solids removal.

Through an intensive testing program, ordered by a federal court, the chemical treatment capacity of the Point Loma plant was investigated. The solids removal and floc formation at the plant can be increased by combining the existing ferric chloride dose with the use of poly-aluminium chloride (PAC). This will remove more colloidal material through better coagulation but will also increase the sludge blanket levels. Increased sludge removal capacity must be installed to maintain a consistent good treatment result.

If the inorganic salts are combined with a cationic polymer, the dosage of ferric chloride and PAC can be reduced by approximately 25 %. If an anionic polymer is dosed prior to the sedimentation tanks the sludge volume will also decrease.

The treatment system with ferric chloride, PAC and cationic polymer at the influent of the plant, and an anionic polymer prior to the sedimentation basins, will improve the solids removal, if the produce sludge is removed reality. This system may lead to removal rates acceptable to the new federal legislation for deep ocean discharge in the United States.

120 000 t/y of dry sludges are produced as a result of the precipitation techniques commonly applied for metals control in the 40 Mm³/y of wastewaters discharged by the tannery industry in Italy. Different approaches for tannery waste management, based on “conservative environmental technologies”, are proposed in the paper, together with a comparative cost-benefit evaluation. Details on a demonstration campaign run using an ion exchange pilot plant (10 m³/d) at a real installation are also given in the paper.

New investigations have shown that stormwater from a separate sewer system is an important pollution factor. The concentration of several parameters is higher than in the treated wastewater and partially exceeds the minimum requirements concerning wastewater discharge into the receiving waters. The load of several parameters in stormwater is equal to wastewater both in the investigated drainage area as well as in the western part of Berlin. With the upgrading of the wastewater treatment plants in Berlin, the significance of stormwater as a factor for the load of receiving waters increases. Therefore it is important to ensure a good treatment of the stormwater. The preliminary tests have shown that for coagulation and flocculation of the very weakly buffered and very soft stormwater, polyaluminium chloride (with a share of cationic flocculant aid), which reacts less acidic, is especially suitable. The optimum dose is 0.06 mmol/l Al³⁺, whereby the pH value has never fallen below 6. In the pilot plant investigations it was possible to work with this constant dose of polyaluminium chloride because the pH value and the acid capacity during the stormwater flow remained relatively unchanged. The best advanced stormwater treatment process is coagulation and flocculation in a pipe designed for both microfloc and macrofloc formation and for separation by filtration. In this case, in particular phosphate, organic compounds, filterable solids and anthropogenic substances, e.g. lead (a high portion of which comes into receiving waters in Berlin by stormwater) and copper could be removed. These results indicate that, in general, this process functions more efficiently than a stormwater tank where particles can only settle.

Stormwater runoff washes off a considerable amount of pollutants from the surfaces of urban catchment areas. Most of the pollutants are transported particlebound and it is well known that solids of a small grain size (suspended solids) serve as their host carrier. In the traditional design, the stormwater sewer of a separate system discharges into an intermediate sedimentation tank at the outfall of the receiving water to reduce the pollutants load. Due to the extended sedimentation time of small particle, this type of stormwater treatment facility is incapable of removing finer, polluted solids fractions. A demonstration project has shown that chemical coagulation and liquid/solid separation by dissolved air flotation results in an efficient removal of pollutants and allows recovery of the natural waters.

Biology does not afford treatment of CSO’s (Combined Sewer Overflow) waters under fair technical conditions. Due to limited upflow velocities during the settling phase, the usual physico-chemical processes do not afford treatment of such high flow rates under fair economic conditions.

The ACHFLO process, using recycled microsand as a ballast to increase the settling speed of flocculated particles, overcomes this economic problem through lowering by a factor of 10 (approx) the lamellar area required to treat a given water flow.

Removal of 80% of TSS, 60% of BOD, 15% of NTK and 80% of P is achieved with a consumption of about 60 g FeCl₃, 0.8 gpolyelectrolyte and 3 g of microsand per m³ of CSO’s water, within a total area of only 0.03 m²/m³ influent per hour.

Plant start-up is achieved within about 15 minutes.

Microsand weighted flocculation appears to be a good way for disposing of the COS’s water from large and medium urban catchments economically.

The factors affecting sludge production rates at wastewater treatment plants are discussed. Methods are presented for the estimation of additional sludge production when chemical precipitation is employed. Sludge production rates at full scale plants employing chemical precipitation are given. When using iron or aluminium salts sludge production increases by 22–35 %. At plants with lime precipitation the sludge production can increase by 100% or more. Experiences with thickening, stabilization and dewatering of sludges from chemical precipitation plants are discussed. With chemical precipitation the release of phosphorus to the water phase during anaerobic digestion is small, otherwise chemical precipitation has no essential impact on sludge treatment. It is concluded that an increase in sludge production is the only negative consequence of chemical precipitation on sludge treatment. Recovery of chemicals may change this in the future.

A full scale demonstration project with thermophilic aerobic digestion (TAD) of primary-chemical sludge was started in May 1993 at Vårnes wastewater treatment plant (7300 p.e.). The project is partly funded by the Norwegian State Pollution Control Authority. The main objective is to document plant performance and characteristics of treated sludge when handling primary-chemical sludge.

The demonstration plant is a standard two-stage facility from the German manufacturer Fuchs Gas- und Wassertechnik operating with a batch (once a day) feeding of sludge. A comprehensive monitoring programme has been established and this paper presents data on volatile solids reduction, pathogen reduction, sludge dewaterability and sludge nutrient content.

A general conclusion so far is that the TAD process investigated here seems to be appropriate for stabilisation and disinfection of primary-chemical sludge, and that plant performance corresponds well with the long-term experiences from German plants treating biological or primary + biological sludges.

Pure water is one of our most important assets today. Therefore there are increasing demands on water treatment. It is our responsibility to return the water to nature in the same condition as we borrow it.

During the start-up phase of an enhanced biological phosphorus removal (EBPR) plant, the fixation mechanisms of phosphorus in the excess sludge were investigated. The major part of the eliminated phosphorus was bound as polyphosphate (poly-P) and was estimated to be some 50 to 70 % of total P. During anaerobic stabilization the poly-P was completely hydrolysed but only a part of the released phosphate remains in solution whereas as much as 60 to 70 % of total P was fixed by physico-chemical mechanisms.

Among the possible counterions for P-precipitation, aluminium, magnesium, and calcium are discussed in detail. In addition to the sludge from the pilot plant, stabilised sludge samples from large EBPR and conventional treatment plants were also investigated. It was found that a major part of the phosphate was associated with aluminium which was demonstrated by means of acidimetric titration. In the pH range of 1 to 3.5, aluminium and phosphate release were strongly correlated with a molar release ratio of 1.45 M A1/M P. The amount of phosphate which was due to aluminium phosphate fixation was calculated as some 35 to 68 % of total P. Besides a precipitation as AIPO₄, adsorption to SiO₂ and/or aluminosilicates and complexation with mono- or polynuclear aluminium are thought to be mainly responsible for the fixation of PO₄.

Although in some sludge samples exceptionally high calcium levels of up to 8.5 % were reached, no indications of calcium phosphate interactions were found.

Studies of sludge conditioning were conducted at five different treatment facilities employing both belt presses and centrifuges. Polymer dose was varied while comparing charge measurements to full-scale dewatering efficiency. Near-zero streaming current correlated with optimum or near-optimum dewaterability under a wide range of conditions. Other assessments of sludge conditioning such as cake and liquid-stream solids, capillary suction time, liquid-stream viscosity, and liquid-stream turbidity led to a more comprehensive hypothesis concerning polymer functioning. Charge neutralization appears to be necessary, but not sufficient, for adequate conditioning: shear effects subsequent to polymer addition determine the polymer dose leading to charge neutralization when phase separation is actually effected. Moreover, such shear itself actually determines floc strength, and the polymer must be capable of maintaining floc structure under these conditions. Since mixing conditions are reasonably constant in installed sludge conditioning facilities, use of the SCD for automatic control of polymer dose appears feasible in principle. Some practical aspects of SCD design and use are also presented that enable polymer dose control to be put into practice.