Dubai has enjoyed one of the most impressive periods of sustained economic growth in the Middle East. This year, Dubai’s gross domestic product growth rate will exceed that of China. The rate of development impacts all sectors, from real estate to finance to tourism. The resulting population growth has fuelled this huge economic expansion, and Dubai stands to more than double its population in the next five years.
Evidently, this explosive growth could not be achieved without an increasing capacity to produce the two major ingredients of any sustainable development: power and water. Moreover, in the Gulf region, the two are intricately linked, since water is primarily produced by desalination systems, which require a tremendous amount of energy to drive distillation or high-pressure membrane installations.
The Dubai Electricity and Water Authority (DEWA) is the government body dedicated to the production of water and power, and ensuring their transmission for the emirate. It is facing the highest growth in electricity demand in the region. An increase of 15 per cent was reported between the 2005 and the 2006 summer peak, making DEWA one of the prime developers of new generation power systems – government plans call for an increase in DEWA’s capacity from about 4 GW to 10 GW by 2012.
The challenges of liquid firing
The Jebel Ali power and water complex in Dubai, adjacent to one of the largest commercial and industrial complexes in the region, is on the front line of reliable and available power supply. DEWA operates a fleet of combustion gas turbines for the emirate with a total capacity of 2.6 GW, a large part of which sits on the Jebel Ali complex. Some 48 km away from Jebel Ali, the six MHI 70ID industrial gas turbines at Al Aweer station were designed primarily to serve peak demand during summer days.
DEWA’s fuel filtration system in action at the Jebel Ali power and water complex
Jebel Ali and Al Aweer gas turbines can be fuelled with either natural gas or diesel. In 2005, DEWA embarked on the challenge to gradually switch from gas to liquid fuelling of most of its turbines. DEWA views this ability to operate continuously on either fuel as a guarantee of reliable supply because it reduces dependency on one type of fuel and increases long-term stability and flexibility of its power plants.
Dual fuel turbines are designed to be able to burn efficiently both liquid and gas fuels. In most practical applications, including at DEWA prior to 2005, liquid fuel is kept as a back-up to the more economical and environment friendly gas-fuelled production. When deciding to use liquid fuel continuously, DEWA faced a set of technical and economical challenges.
Fuel of choice
Natural gas is the fuel of choice for diverse reasons, some fluctuating and related to market forces, but some are also technical, and unchanging. One key issue faced by DEWA was the treatment of the liquid fuel prior to injection in the turbine combustion system.
Ensuring consistent quality of the fuel is both an operational and technical challenge. Liquid fuel has the disadvantage of being both more variable in terms of cleanliness level/calorific value and potentially more contaminated than its gas counterpart. Equally important, liquid fuel can contain a much higher quantity of water, often in the form of small water droplets. In order to avoid problems generated by solid and moisture contamination in liquid fuel, typical treatment systems comprise both particulate filters, designed to remove solid contaminants of a given size and fuel coalescers, able to separate and segregate free water from the fuel stream.
The efficiency of such a coalescer is largely dependent on their pore size, or the size of the channels through which microscopic droplets travel and are forced to coalesce into larger drops to then be separated. Some of the best coalescers are able to remove water down to the fuel saturation level, meaning no droplets can escape separation, regardless of how small they might be.
This in turn means that the coalescer channels (or pores) have to be very small, typically under 20 microns in size. In theory, coalescers can last forever because the separation method is not one of capture but one of coalescence. In reality of course, coalescers are subjected to the various other forms of contamination present in liquid fuel, and typically reach the end of their life when their pores become filled, not with water but with particles and gelatinous contaminants. Hence the filtration is always upstream of coalescence, in order to protect not only the turbine, but the coalescers themselves. In short, a high efficiency coalescer require tight upstream protection.
The first DEWA turbines to experience extended diesel firing were the six Mitsubishi 701D of the Al-Aweer H station. When intermittently operating their turbines with liquid fuel, DEWA operators were reporting an average lifetime of the installed coalescers of just two or three days. This situation can be managed for a few days of liquid combustion, but DEWA engineers understood that this was unsustainable in continuous mode, and hence decided to embark on a programme to improve the reliability, availability and cost effectiveness of their fuel treatment system.
After reviewing existing literature and current fuel treatment practices at other DEWA plants, a study concluded that pre-filters of suitable size and specifications were needed. DEWA approached Pall Corporation and Alphamed, Pall’s representative in the UAE, with the objective of not only ensuring optimum fuel cleanliness entering the turbine, but also better protection and longer life for the coalescers installed in the system.
A step-by-step approach
The first step was to filter bulk liquid fuel held in storage tanks on-site, historically kept as standby. The goal was twofold: ensure proper start-up of the turbines with no risk of contamination from settled diesel, and evaluate the expected lifetime of the new filtration system using a fuel as close as possible to local market purity standards. This initial tank recirculation was achieved using an old stock of coalescer elements.
The second step was to carry out a test that would closely replicate actual operational conditions without having to fire the turbine with very expensive liquid fuel. DEWA engineers designed an innovative real-time inline test by installing a pre-filter skid from Pall upstream of existing coalescers in one of Al Aweer Mitsubishi type 70ID combustion turbines.
New piping and valving allowed the diversion of diesel fuel back to storage tanks without firing. This arrangement allowed the test to be carried out for an extended period of time, during the low demand winter season where gas supply was sufficient. This in itself resulted in multi-million dollar savings for DEWA from recycled liquid fuel.
After about 45 days of treatment, a laboratory analysis of the filter elements confirmed both the efficiency of the filter in removing particles and gels, and the integrity of its media and construction. The filter, designated as Pall Ultipleat Profile element, benefits from crescent shape construction, a pleat design that allows maximum flux rate and longer life by distributing the flow evenly through the many layers of the filter media. Subsequently, a total of four holding tanks were treated at a flow rate of 70 m3 per hour through one set of 24 filter elements, each treatment during an average five days.
When used on the gas turbine supply line, the filter elements were subjected to six bar working pressure at a flow rate of 70 m3 per hour. As expected, the filter elements reached a pressure drop of 1 bar after six weeks of treatment. DEWA engineers decided to set the replacement frequency for the elements at one month or one bar, whichever is reached first.
Real time operation
Based on the set of kidney loop treatments and the six-week live trial on the first Mitsubishi turbine, the decision was made to replicate the filtration skids upstream of the coalescer vessels on the five remaining 701Ds and three Siemens/Ansaldo V94.2 units of Al Aweer H station.
Since then, the filter elements average a lifetime of 21 days. It is interesting to note that the variance can be quite large, from two weeks to two months depending on the diesel level of cleanliness, showing once again the technical challenge posed by a fuel source that is variable in nature, depending on origin, holding time and transportation methods.
In continuous operation on the turbines, the improvement in filtration characteristics has allowed the coalescers to last up to 30 days, or more than a ten-fold increase in lifetime. The extension of coalescer element lifetime has allowed DEWA not only to reduce the numbers of parts held in stock, but also to reduce the maintenance time associated with coalescer changes, an operation that necessitates a dedicated team to perform. Going through this operation every 20 or 30 days, instead of every two to three days, is a tremendous achievement that also impacts favourably on the environment and operators and plant safety (less human intervention on the system, fewer liquid fuel leaks, etc).
Today, a total of nine filtration systems are installed on gas turbines operated by DEWA, and up to six more could be installed in the near future. Based on almost a year’s experience, the benefits of better coalescer protection are evident. But some work remains to be done, notably defining whether the process of coalescence itself could be optimized.
DEWA and Pall are considering the possibility of field experiments to evaluate the actual efficiency of the installed coalescers, and whether it, besides lifetime, can also be improved. DEWA operational field experience is that very little water is extracted at the bottom of the coalescer. This could be due to the fact that the diesel fuel is delivered and supplied dry – that is its water content is below saturation at the operational condition of the coalescers – or simply that the coalescers are not optimized.
Coalescer efficiency is in part dependent on the pore size inside its structure. Many other factors, like materials, chemical treatment of the fibres, have a tremendous impact on the ability to coalesce very small water droplets out of the fuel stream.
If the coalescers lack the proper structure, coalesced droplets of water are simply re-entrained in the fuel stream. Some additives like surfactants are also known to disarm coalescers, which then become basic filters. In this case, coalescers will plug without performing their dedicated function.
DEWA has put in place a design allowing its combustion turbines to perform continuously on liquid fuel without the risk of solid contamination fouling the coalescers, or worse entering the combustion side of the turbine. By designing a filter element that is both very efficient in the particle size range of the coalescer, and itself long lasting, Pall and Alphamed have shown the impact of better protection of liquid coalescers on the operation and maintenance of DEWA’s combustion turbines.
Pall and DEWA are expected to continue to further optimize the fuel supply line, especially through the evaluation of current coalescer performance. After saving the life of the coalescers, it is now important to demand the most from them.