Achieving consistent low energy consumption in sea water RO based desalination plants using Lowatt® technology

Abstract

Biofouling remains the single most important factor that increases energy consumption as time progresses in the operation of the Reverse Osmosis (RO) membrane system. One of the challenges with plant operation is that once a plant has been designed for certain energy consumption, it does not remain steady over a period. This is mainly due to biofouling and the inability to clean the membrane efficiently.

In an effort to maintain healthy operational efficiency in terms of water production and energy consumption, differential pressure across the membranes should remain constant and close to the start-up conditions for multiple years. To achieve sustainable lower energy consumption, it is important to keep the membranes clean and then adopt an effective cleaning at the very initial phase of biofouling formation before it becomes a permanent problem. This paper describes the combination of ultrafiltration and bio-foulant removers, at a low flux operation, reduces the severity of the biofouling and results in lower energy consumption.

The LoWatt® process offers several advantages by reducing the severity of biofouling and sustained plant operation is possible at lower energy consumption with high plant availability.

INTRODUCTION

Water desalination is growing to meet industrial and drinking water demands worldwide. Although both thermal desalination (multi effect distillation – MED, and multistage flash evaporation – MSF) and membranes-based seawater reverse osmosis (SWRO) processes are used in these plants, SWRO has grown predominantly over the last 15-20 years. SWRO has become very cost effective and efficient in terms of energy consumption compared to where the technology was a few years ago [1]. Most desalination systems incorporate one or the other type of energy recovery devices with a view to optimizing the energy consumption, as energy cost constitutes close to 60% of the operating costs in a SWRO plant [2]. Most designs focus on hardware approaches – with incorporation of energy recovery turbines or VFDs on certain drives – wherein the energy cost savings are calculated to offset the investment in equipment and provide a significant saving over a period.

Though this approach does limit the energy consumption at the design stage, it does little to sustain the energy savings over a period of time, as biofouling starts to build up.

Bacterial deactivator is a serious and recurrent problem in RO plants as it reduces productivity of water, increases the differential pressure, and increases power consumption. This problem is compounded in plants where there are open intakes and where seawater temperature increases during the summer [3]. Chlorine treatment makes this worse due to formation of oxidized products, which provide potent feed for the residual bacteria on the membrane surface where they are rejected along with the bacteria after the de-chlorination process. It has been recognized that chlorination cannot be considered as a sustainable process option to control biofouling. The balance bacteria left after chlorination multiply much faster after de-chlorination with the potent nutrients as their food [4]. Moreover, chlorinated organic products may be undesirable due to the formation of carcinogens. Therefore, alternative techniques to control, minimize or eliminate biofouling are essential.

Biofouling results in a sticky EPS (Extracellular Polymeric Substances) layer on the surface of membrane. It has good shear strength, hence hard to remove. The onset of biofouling first increases the feed pressure, which results in increase of the energy consumption. As the fouling continues, it reduces the water production. To maintain design productivity, the feed pressure must increase, which again increases the power consumption. As this process continues the membranes must be taken for a long and multi-step chemical cleaning process. Simultaneously, the differential pressure across the membranes increases and it becomes difficult to clean the membranes and regain the original performance. Aggressive cleaning of the membrane may result in damage, which may become irreversible with time and ultimately shorten the membrane life.

Figure 1 represents a typical behavior of such a system, where a continuous creep of feed pressure is experienced. It may be noted that this can persist even at lower silt density Index (SDI) range <3 as required by the membrane manufactures.

LOW ENERGY RO (LoWatt®) We present an energy efficient RO desalination, which focuses on achieving Low energy consumption by reducing the biofouling in the membrane integrated with a cleaning methodology. This prevents build-up of any residual bio film on the membrane surface. To achieve sustainable lower energy consumption, it is important to ensure the membranes do not foul, and the differential pressure does not increase. Also, a cleaning methodology is available to clean membranes at the very initial phase of biofouling formation before it impacts differential pressure and before any fouling becomes permanent and starts impacting plant performance in terms of water production, power consumption and product quality. This is made possible by the following innovative process approach:

1.0 The pretreatment is done with the ultrafiltration membranes, which gives more than 6 log reduction of bacteria and 1-2 log reduction of virus. The Permeate of ultrafiltration provides an SDI of less than 3 and very often between 1-2. The UF is able to remove the majority of suspended particles including those which are colloidal in nature, and it also removes some bio foulants. But it is not able to eliminate all the organic contaminants, which participate in fouling on membranes. To calibrate the performance of the UF it is important that inlet water has a turbidity of less than 5 NTU even during upset conditions and any treatment on the upstream of UF is designed to achieve these parameters based on water analysis and site conditions. This ensures UF treated water quality remains around 0.06-0.08 NTU turbidity and SDI values of less than 3. This also ensures the downstream system is protected from any loads of colloidal particles.

2.0 Along with ultrafiltration pretreatment, bacteria deactivator unit is also provided that deactivate the bacteria with the help of electrochemical process and coagulates the majority of organics like humic acids, polysaccharides, proteins, amino acids, carbohydrates, bacteria and other potential contaminants which aid biofouling. Coagulated organics can be then easily filtered through ultrafiltration.

The bacteria deactivator device interferes with the chemical conditioning process of the membrane in controlling biofilm development. When a clean membrane surface is exposed to seawater or natural water, polymeric natural organic compounds are adsorbed. This chemical conditioning imparts a negative charge to the fouled substrate. The conditioned surface can now concentrate lower molecular weight substances used as a food supply for bacteria. The bacteria deactivator system does not allow the carryover of the negatively charged organics and therefore disrupting the process of bio film formation.

Ultrafiltration membranes without bacteria deactivator device does not give much reduction in TOC value in treated water. Whereas with bacteria deactivator, ultrafiltration membrane delivers large quantities of treated water with much more reduced turbidities and SDI while removing a majority of potential contaminants that can cause fouling. The typical SDI value at the outlet of ultrafiltration is less than 1 and typically close to 0.6-0.8. The process highlights the importance of presence of bio-foulants along with UF, which is critical for eliminating or minimizing bio film formation at reduced flux of RO. There are multiple options of bio-foulant removal, which operate under a wide range of Total dissolved solids (TDS) and provide Total organic carbon (TOC) reduction of at least 40-60% on an overall basis and remove the majority of the negatively charged TOC. The bacteria deactivator can include ion exchange materials, positively charged media or electro chemical or electrode-based methods.

3.0 The system Design and plant operation is done at a lower flux of around 6-8 GFD based on feed water quality, permeate quality requirement and temperature range. This is achieved through a low flux Reverse osmosis (RO) process.

The flux can be marginally increased for lower (TDS) or low fouling waters. This flux is optimum for energy consumption as there is not much improvement in energy consumption if the flux is further reduced. Any further reduction in flux will cause deterioration in permeate quality. This flux reduces the concentration of bacteria and nutrients over the membrane surface and reduces the differential pressure to a minimum. Moreover, at this reduced flux, operating pressures reduce significantly by a minimum of 10-20%. Also, there is minimum variation in the difference in operating pressure with variation of feed water temperatures. When the design flux is higher as per the conventional process, there is significant variation of operating pressures at minimum and maximum temperature. This requires sophisticated controls to adjust the pressures, but these still result in loss of energy when the actual temperatures are higher than design. Alternatively, speed control devices have to be installed to adjust pump speed for changes in water temperatures, which still result in some loss of energy and make the system complex and expensive. Operation at low flux design avoids this complication and reduces energy consumption by 20%. For example, for 35000 PPM TDS, if the system is designed at 9-10 GFD the power consumption is around 2 KWH/M3 for the RO pump and energy recovery system. But if the same system is designed at 6 GFD the power consumption reduces to 1.7 KWH/M3 (reference Figure 4) and reduces feed pressure from 55kg/cm2 to 46 kg/cm2 (reference Figure 3). At this level variation in pressure due to feed water temperature within a wide range of 25- 40 degrees C is only 0.5-0.7 kg/cm2 (reference Figure 5) for different types of membranes. This also provides product TDS within acceptable limits even at the highest possible temperature (reference Figure 6).

The energy consumption has been calculated based on 85-86% efficiency of pumps and more than 96% efficiency of motors. This data is more or less consistent for different membrane models available from different membrane manufacturers in the market and their difference, if any, is very small. It is evident from these studies that at these levels of flux the energy consumption is most optimum and can handle a wide range of temperature with minimum variation in power and also provides acceptable range of permeate TDS. At this level of flux, the bio film formation is reduced to insignificant levels especially when it is pretreated with UF and the bio-foulants removal unit as mentioned above. This ensures that the energy consumption design is low to start with and remains low on a sustained basis due to reduced or insignificant biofouling. Over a period of a day’s operation, the increase of differential pressure is less than 0.1Kg/cm2, and more often less than any detection limits. Also due to reduced driving pressure across the membrane, whatever fouling happens is not firmly attached to membrane surface and therefore can be easily removed under mild cleaning conditions. With certain precautions taken in the pretreatment as described, the residual foulants are not able to adhere to the membrane surface. Some of these concepts are similar for surface water RO plants, including some low TDS waters, where severe fouling happens on reverse osmosis and energy consumption increases and water production eventually drops. It has been seen that biofouling alone can increase the differential pressures across RO stages to more than 4-5 kg/cm2, which results in he loss of energy. This may happen even if the pretreatment includes a UF system. This can be easily mitigated by optimizing flux, calibrating and regulating pretreatment as described in points 1 and 2 above and stopping the buildup of biofouling as mentioned below in point 4.

4.0 To further augment the process described above and to overcome any biofouling right before it initiates, a unique methodology of cleaning is devised based on natural osmotic pressure differential between the reject and permeate water. When the system or part of the system is stopped with a continued regulated flow in the feed side, which allows the reject water to remain in the feed side, there is a steady flow of water from the permeate side to the feed side. The permeate flow continues to the feed side due to concentration differential. The concentration differential is maintained by makeup reject water flow through a recirculation system. If this process is allowed to continue for 10-15 minutes, any bio film is dislodged from the membrane surface. As the plant has been designed at lower flux and also the feed water has been filtered through UF with bacteria deactivator unit, the buildup of any bio film and pressure drop is reduced and can be controlled through the osmotic cleaning. This process should be controlled through regulated flows and concentration on both feed and permeate side using plant produced reject and permeate water. The permeate back flow under these conditions is purely a function of concentration gradient and pressure drop built in the membranes due to fouling. The feed side flow is adjusted by circulation of brine to overcome dilution due to permeate entry and also maintain dynamic conditions in the feed side. It is possible to maintain clean membrane pressure drop conditions by using this cleaning technique and prevent any increase in feed pressure or membrane differential pressure. The loose debris can be then flushed out with pretreated seawater rinsing at a higher velocity. This cleaning methodology is based on a concept that bio film formation should be removed as fast as it is formed or prevented from building up. It can be achieved by shorter cleaning cycles of 10-15 minutes done frequently or based on predetermined differential pressure increase over start up conditions. Normally the differential pressure builds up from 0.1 kg/cm2 per day to 0.3kg/cm2 a day over 24 hours of operation depending on site conditions and plant design. This process will typically not allow any buildup of differential pressure, and the membrane will operate at clean membrane conditions. This process does not use any cleaning chemicals on a daily basis but uses the brine generated in the reject of SWRO or BWRO plants. The option of adjusting brine concentration can be exercised to control the effectiveness of the cleaning process. The cleaning processes will be carried out with osmotic cleaning as a feature for LoWatt® technology. The osmotic cleaning of SWRO train is carried out without affecting the production. The same is done by an automated system, which may include a biosensor. The input based on differential pressure, or a biosensor shall trigger the cleaning procedure of individual banks with a set of SWRO tubes, which will get isolated from the train [5]. The rest of the SWRO train will continue the water production.

Performance trials on a reverse osmosis system with biologically active water.

Performance Trials 1: With UF as a pretreatment step alone.

To benchmark the base performance, a Reverse Osmosis (RO) unit operates for 17 months on surface water having TOC level of 5 – 10 ppm without any bio foulant removal unit at the upstream of the RO unit.

This source of water was selected due to its previous history of biofouling for several years. Based on the original plant design, the surface water was passed through an Ultrafiltration (UF) unit before feeding into an RO unit and the silt density index (SDI) was maintained below 5 and most of the time below 3. The RO unit pressure drop was monitored, and its results are as shown in Figure 8.

During the 17 months of operation, the RO unit was cleaned seven times to keep the pressure drop constant. The RO unit’s average service cycle time was approximately 700 hours, and chemical cleaning was necessary to maintain pressure drop, product quality, and energy usage. Table 1 shows the running hours of this RO unit throughout several servicing cycles. The running hours were gradually altered such that, after each chemical cleaning, the original starting pressure drop circumstances could be restored. During this operation, it was very clear that even with the UF pre-treatment process, pressure drop increase was evident within days and sometimes within hours during the rainy seasons and after a very elaborate cleaning process the initial pressure was not restored.

  Table 1: RO Unit Operating Hours vs. Service Cycle

Performance trials 2: on a bio-foulant removal unit

In this experiment, a bio-foulant removal unit (bacteria deactivator) was installed with UF. The TOC and turbidity values across the system with and without bacteria deactivator unit were monitored and compared. The results of TOC and Turbidity are shown in Table 2.

Data above in Table 2 clearly indicates that the bacteria deactivator unit improves the TOC removal efficiency of UF by around 40 – 60% and outlet turbidity of water was always around 0.060 NTU. This directly helps in maintaining the SDI level below 3 in the RO unit and mostly in between 1–2, minimizing the biofouling in the RO unit.

Performance trials 3: Combined pre-treatment of UF followed by bio-foulant removal unit

In another set of experiments, the same RO unit was operated in two conditions as shown in Figure 10-A and 10-B with the inclusion of the bacteria deactivator unit at the upstream of the UF/RO unit. Performance of the RO Unit in these two conditions is summarized in Figure 11.

In Condition-1, the RO Unit was operated for nine months (Approx. 1400 hours) and its effect was clearly observed with respect to longer service cycle length as compared to the service cycle lengths of trial-1. During this operation it was observed that for more than 3 months there was a very insignificant increase in differential pressure but once it started increasing, gradually subsequent fouling rate started accelerating and progressively started increasing. Even though the bacteria deactivator unit minimized the pressure drop rise and bio-fouling in the RO unit, still the pressure drop gradually increased over a period of six months and the main reason for this is the gradual deposition of fine bio film on the RO membrane surface day by day. The intensity of biofouling was very low as indicated by longer service length.

In Condition-2, after the normal chemical cleaning of the RO unit and bringing back its pressure drop to a normal level (3.8 kg/cm2), a natural osmotic cleaning process was implemented and every day one natural osmotic cleaning cycle was performed on the RO unit by RO Reject water for 10 – 15 minutes. The impact of natural osmotic cleaning was clearly observed, and the pressure drop was unchanged at 3.8 kg/cm2 for next 1000 hours of operation. Due to the unchanged pressure drop of the RO Unit, its energy consumption remained the same, no increase was observed. During this time, no increase of differential pressure was seen. It became clear at this stage that with proper feed conditions of UF and bacteria deactivator unit and proactive Osmotic cleaning, virtually clean membrane conditions can be maintained, which means no biofouling and no increase in energy.

Performance trials 4: with Lowatt® process

In this study, the RO plant was operated with a new set of membranes with LoWatt® Process that includes a UF system with bacteria deactivator unit at the upstream of RO unit and a periodic osmotic cleaning system. It was operated for 20 months (approximately 5000 hours) with the same water source used in trial-1 and the expected results were observed without any increase of differential pressure at sustained levels of power consumption. Figure 12: RO Unit performance graph when operated as per LoWatt® Process.

 The Observations

• The differential pressure drop across the membrane did not increase (see Figure12) and remained constant.

• The RO system with LoWatt® process was not

• once cleaned and clearly differentiates its advantages over conventional RO where the plant was cleaned seven times during same duration and similar conditions. (see Figure 13).

No changes or increase in power consumption were experienced with LoWatt® process. It remained mostly constant throughout the period.

3.5 Advantages of LoWatt® Technology

1. Lower energy RO technology, especially for surface water and SWRO applications

2. Optimum flux design for lowest initial energy consumptions

3. Initial and then sustained low energy operation

4. Energy consumptions for a SWRO system ranging from 2.8-3.0 kwh/m3

5. Lower biofouling due to lower concentration of bacteria and food per unit area of membrane, lower membrane replacement rate

6. Online natural cleaning and minimum chemicals

7. Predictive diagnostics and cleaning option- based on Bio senor [5]

8. Focus on the root cause of biofouling, i.e. bacteria and organics

9. Reduced chemical consumption

Conclusion:

The above studies clearly demonstrate that LoWatt® presents itself as a significant development in lowering the cost of desalinated water by as much as 25% than would be possible with best known energy recovery approaches available today. In addition to the lower initial power consumption, the system is also able to sustain the design power consumption without compromising on product quality, thereby providing a predictable life cycle cost for a desalination system. From a plant manager’s perspective, this means more system availability and reliable operations without expensive and time-consuming chemical cleaning regimes.