1 District Cooling as an Energy and Economically Efficient Urban Utility Its Implementation at Marina Bay Business District in Singapore Tey Peng Kee, Singapore District Cooling Pte Ltd Abstract-- In a resource-constrained country like Singapore, energy conservation and efficiency are strategic measures to reduce carbon footprint. As air-conditioning accounts for major energy usage in commercial buildings, chiller plant efficiency has attracted considerable interest and attention. District cooling is an energy efficient outsourcing alternative to in-building chilled water production. This paper presents as a case study the implementation of district cooling system at Marina Bay, a new business district in Singapore. In addition to the technical features of the system and attributes of the new utility service, the paper describes the regulatory framework and commercial arrangement adopted. The paper also provides a comparative discussion on how a district cooling system could effectively elevate the energy efficiency for airconditioning in a commercial district in a cost-efficient manner. Index Terms Air conditioning, chiller plant, district cooling, energy efficiency, utility service. I. INTRODUCTION Singapore relies completely on import for its energy needs. Energy conservation and efficiency are key strategic responses to address growing concerns on global warming. For a city in the tropics, air-conditioning is an essential service for commercial buildings. In Singapore, about 70% of electricity usage in commercial buildings is related to air-conditioning and two third of which in chiller plants. Chiller plant efficiency has accordingly become a key area of attention in initiatives to improve the energy efficiency of commercial buildings. District cooling is an outsourcing alternative to inbuilding chilled water production for air-conditioning. It is well suited for commercial districts with high cooling load density. It raises the energy efficiency related to airconditioning more effectively on a global scale. A large-scale district cooling system has been successfully implemented for Marina Bay Business District in Singapore (Fig. 1). Fig 1. Marina Bay Business District (an artist s impression) II. BRIEF HISTORY In mid 1990 s, Urban Redevelopment Authority (URA), the town planning authority in Singapore, conducted feasibility study and planning for the construction of common services tunnels in the new business district for accommodating utilities lines. As part of the feasibility study, district cooling was identified as an urban utility for the new business district. Provision was made in the design of the common services tunnels for accommodating chilled water pipes. Singapore Power 1 and Dalkia 2 conducted a feasibility study and confirmed the commercial viability of the new utility service. They went on to form Singapore District Cooling (SDC) as a joint-venture to implement the pilot system. The first district cooling plant was completed and commercial operation commenced in May 2006. III. DESCRIPTION OF THE SYSTEM The original Master Plan envisaged a service area with over 8,000,000 sq m in gross floor area. The cooling load was estimated to be about 900 MW r, which could be served by five district cooling plants. In order to optimize land use, the district cooling plants are to be co-located within selected large-scale developments. The District Cooling System now comprises two chilled water production plants which are interconnected by a piping network (Fig 2). 1 Singapore Power Group is a leading energy utility group with headquarters in Singapore. It owns and operates electricity and gas transmission and distribution businesses in Singapore and Australia. 2 Dalkia is the energy division of Veolia Environment, a French utility group in environmental services. It is a leading provider of energy services across Europe, North America and Asia.
2 Plant 1 Piping Network MW r by end 2010. Cooling demand is projected to grow rapidly in the next few years as new developments come on stream in the new business district. Marina Bay Financial Ctr Marina Bay Sands Plant 2 E. Hot Water Supplies As a green initiative and also a business enhancement, hot water is generated from waste heat in the condensing water circuit at Plant 2. Heat pumps (4 x 900 kw r ) provide hot water source for the hotel and food establishments in the integrated resort. Fig 2. Marina Bay District Cooling System A. District Cooling Plant No 1 (Plant 1) Plant 1 is located at One Raffles Quay, a premium office complex completed in 2006 on the site of first land sale for Marina Bay which was successfully concluded in 2002. The machine room occupies the basement of the complex and cooling towers sit on the podium roof between two office towers. Plant 1 has a design capacity of 157 MW r. Phase 1 of the plant, with 57MW r production capacity, was commissioned in May 2006. In October 2009, additional 40MW r production capacity was installed under Phase 2 of the plant. Currently, Plant 1 has an installed capacity of 97MWr comprising the following key equipment: 1 x 3 MW r water chiller; 2 x 7 MW r water chiller; 2 x 10 MW r water chiller; 3 x 10 MW r brine chiller; 3 x ice thermal storage system, each capable of 10MW r discharge rate; and 18 x 5 MW r cooling tower B. District Cooling Plant No 2 (Plant 2) Plant 2 is located under the Bayfront Avenue forming part of the land site for the new integrated resort developed by Marina Bay Sands. Plant 2 is designed for 180MW r production capacity. In May 2010, Phase 1 of the plant was completed with 60MW r installed capacity comprising the following key equipment: 6 x 10 MW r water chiller; and 4 x 20 MW r cooling towers. C. Chilled Water Piping Network Interlinking the two plants is a chilled water piping network which is installed in the common services tunnels. The trunk section of network comprises pipes of 1.5m diameter. Branch pipes connect buildings to be served to the network. The pipes are insulated with foamglass with outer galvanized steel sheet cladding. As at May 2010, the network pipes total 5 km in length. D. Cooling Demand The contracted supply capacity will amount to 113 IV. COMMERCIAL ARRANGEMENT WITH CUSTOMERS i) Technical Scheme Chilled water supply to a customer is provided via heat exchangers as a connection interface in an Intake Station (or energy transfer station) located within the customer s development. The typical connection scheme is shown in Fig. 3. Fig 3. Typical Intake Station Schematic The building space for the Intake Station is provided and maintained by the customer. The Intake Station also accommodates the downstream pumps for distribution of chilled water within the development. The costs of the service connection facilities, which include the heat exchangers, metering / control equipment and connection pipes to the Chilled Water Piping Network, are borne directly by the customer but the installation and maintenance of the facilities are undertaken by SDC. SDC bears the costs of upstream piping network and the district cooling plants as infrastructure costs which are translated into recurring monthly Contract Capacity Charges levied on the customers. ii) Performance Based Service Agreement Chilled water supply temperature is regulated at 6.0ºC +0.5ºC. The customer is required to adopt variable flow design for its downstream reticulation so as to achieve a return temperature higher than 14ºC. If the hourly average supply temperature exceeds 6.5ºC, SDC pays a rebate that is twice the equivalent hourly rate for Contract Capacity Charge. Similarly, if the monthly average return temperature falls below 14ºC, the customer pays a surcharge on the Usage Charge.
3 iii) Tariffs The chilled water tariffs are regulated by Energy Market Authority (EMA) and the tariff rates are reviewed at half-yearly intervals. There are five components as follows:- a) Contract Capacity Charge This is based on the supply capacity requirement declared by the customer. The rate is benchmarked to the fixed costs of operating in-building chiller plants. The benchmark parameters were established by the Energy Market Authority through an independent survey of chiller plants in ten commercial buildings. b) Usage Charge The payment is based on metered energy consumption. The rate is pegged to prevailing prices for electricity and water which are variable inputs for chilled water production. c) Capacity Overrun Charge This component provides flexibility to a customer to address his ad-hoc requirement for higher cooling capacity and allows him to declare his supply capacity requirement to correspond to his sustainable demand. The daily rate for Capacity Overrun Charge is one-tenth of the monthly rate for Contract Capacity Charge. d) Return Temperature Adjustment Low return temperature from customers (commonly known as Low Delta-T Syndrome ) degrades the system performance. It results in additional pumping in the chilled water supply network. It also renders it difficult to load chillers to their design capacities, thereby adversely affecting their energy efficiency. To encourage good operation and maintenance practice at customer s end, there is an upward adjustment of the Usage Charge by 3% for each degree C of the monthly average return temperature below 14ºC. c) Supply Deficiency Rebate SDC is committed to high supply quality. A rebate equivalent to twice the amount for the corresponding Contract Capacity Charge is paid to the customer when the average supply temperature for any hourly interval fails to meet the specifications. V. REGULATORY FRAMEWORK District cooling, as an infrastructure system, derives its economic advantages primarily from its large-scale operation when compared to in-building chiller plants. On the other hand, developments subscribing to district cooling service, as in the case of any other public utilities, have little flexibility to discontinue the subscription of the service in future. The concerns about the lack of control over the supply quality and pricing of the new utility service present major impediments for a new district cooling system to attain the critical mass of demand when development owners have the option to continue building their own in-building plants. The authorities in Singapore accepted the submission for district cooling to be made a mandated utility service in order to mitigate the start-up commercial risks. Accordingly, the District Cooling Act was legislated in 2002 to provide the necessary regulatory framework. The legislation, administered by the Energy Market Authority of Singapore, requires that the new utility service be priced at a level no higher than the equivalent costs of chilled water production in conventional inbuilding plants employing similar technology. Over time, the district cooling operator is allowed to earn a baseline return based on its invested assets. When the operator has recovered its start-up losses after achieving the critical mass of demand for efficient operation, any efficiency gain above the baseline return shall be shared equally by the operator and customers. Customers are thus assured of long-term savings and the start-up demand risk for the operator is also mitigated. VI. COMPARATIVE ADVANTAGES OF DISTRICT COOLING Compared to in-building independent chiller plants, a district cooling system is superior in terms of asset efficiency, energy efficiency and service level. A. Asset Efficiency To ensure service continuity, spare machines are always needed in in-building plants to cater for maintenance and equipment failure situations. Often for consideration of ensuring adequate cooling capacity, the design engineer may also provide additional design margin. Comparing the installed capacity to the actual peak cooling demand, such capacity margin could range from 50% to 300%. Once installed, the spare capacity is effectively a stranded asset. Such asset stranding does not occur for a district cooling system where the system expansion takes the actual cooling demand into consideration. Any unused capacity at any point in time can be used to serve new developments in due course. In addition, due to diversity of cooling demand by different buildings, the system capacity required to provide supplies to a group of buildings is always lower than the aggregate capacities that need to be installed for individual buildings. There is thus better asset efficiency in chilled water production facilities through the aggregation of cooling demand to be served by centralized plants.
4 However, a district cooling system necessitates a piping network. Except for high load density area, the investment in the piping network could erode the asset efficiency of chilled water production. B. Energy Efficiency While larger and more efficient machines can be deployed in a district cooling system, this is not a distinctive advantage. Higher energy efficiency rating can normally be achieved, even in in-building chiller plants, with higher investment. Upon attaining its critical mass of demand, a district cooling system is inherently more energy efficient than independent in-building plants primarily for two major reasons:- - High loading level for machines close to their design capacity; and - Professional attention to operation and maintenance of the system. As a typical in-building plant has few chillers, there are practical difficulties to match the capacity of online chillers to the actual demand all the time. The cooling demand of a building varies, inter-alia, with the type of usage, occupancy, time-of-the-day and weather conditions. In addition, in order to ensure adequate cooling capacity, it is not uncommon for the design engineer to add a design margin to the cooling load assumption which typically already represents the worstcase cooling demand. Many in-building chiller plants also suffer from the Low Delta-T syndrome. Due to all these factors, chillers in independent in-building plants typically operate at a low loading level most of the time. The energy performance accordingly falls short of the design intention. As there are many more machines needed to meet the aggregate demand of many buildings, the chillers in a district cooling system can be scheduled to run based on the actual cooling demand in a manner similar to electricity generation scheduling in a power system. Thermal storage facilities in a district cooling system also provide flexibility to delay or avoid the starting of an additional chiller when the demand marginally exceeds the aggregate capacity of the online machines. Hence, the loading level can be maintained high, closer to the design capacity and efficiency of the machines. Very often, skilled and experienced professionals are also not available or deployed for operation and maintenance activities of standalone in-building chiller plants, leading to degradation in energy efficiency. On the other hand, the operational efficiency of a district cooling system critically affects the profitability of the utility business. The operation of a district cooling system is closely supervised by more sophisticated control systems and professional attention is accorded to its maintenance on an on-going basis. Table 1 summarises the energy performance of the district cooling system at Marina Bay. Table 1 : System Energy Efficiency in June 2010 Direct chilled water production 0.71 kw e h per RT-Hr Ice production 1.03 kw e h per RT-Hr Network pumping 0.04 kw e h per RT-Hr System overall 0.80 kw e h per RT-Hr The System Overall measure is computed on the basis of electricity input to the district cooling system compared to the cooling energy sold. With growth in the system demand and continued fine-tuning effort for system operations, the energy efficiency is expected to improve further. Notwithstanding the additional network pumping and lower conversion efficiency for thermal storage system (TES), the current System Overall energy efficiency indicator at 0.80 kw e h per RT-Hr is more superior to the performance of most in-building chiller plants in Singapore, for which the average system efficiency is about 1.0 kw e h per RT-Hr. The energy efficiency margin is actually higher if an adjustment is applied to account for the difference in supply temperature of the chilled water from the district cooling system (6.0ºC) and in-building plants (typically in the range of 6.5-8.0ºC). Table 1 shows that generation of cooling energy through an ice TES uses some 40%-45% more electricity input than direct chilled water production. Critics of district cooling system often cite the deployment of ice TES as out of line with the global effort of reducing greenhouse gas emission. This is a misguided perception. The use of TES effectively shifts some electricity load to off-peak night hours for cooling demand during the day hours. Incremental electricity generation in the power system at off-peak requires less source fuel compared to on-peak period when less efficient generators are required to come online. In Singapore, the efficiency of power generation ranges from below 30% for open-cycle gas turbines to some 38% for steam plants and over 50% for combined cycle plants. Hence, from the perspective of source fuel usage, the higher electricity input required by ice TES but at off-peak hours does not necessarily amount to more greenhouse gas emission per unit cooling energy produced. The benefits for the deployment of ice TES in the district cooling system at Marina Bay include the following: a) Although the spread between the peak and off-peak electricity prices is small in Singapore, lower overall per-unit energy cost can still be achieved through higher load factor for electricity input to spread the electricity capacity charges and opportunistic reduction
5 of electricity supply intake during periods of price hikes in the wholesale electricity market. b) The limited roof space available for cooling towers can be optimally utilized for more cooling output capacity from the district cooling plants. c) With TES, the cooling towers, plant infrastructure such as electrical system and the fewer number of chillers are worked harder to provide the peak cooling capacity. This is advantageous from the perspective of total investment costs. d) The chilled water supply reliability can be enhanced with TES, particularly at start-up period of the system. e) The TES also functions as a regulating source of chilled water supply to facilitate online chillers operating at full load for energy efficiency. There is indeed a need for caution when comparing the kw per RT numbers for chiller plants. While they serve to provide a general indication, they are not a comprehensive indicator for evaluating the energy performance without concurrently taking into consideration of the technology, locations and operating conditions of the plants. C. Higher Service Level District cooling is a utility service like electricity. It is superior to chilled water supplies from in-building plants for these attributes: - round-the-clock availability; - well regulated supply temperature; - supply as-demanded at any quantity within the contracted capacity; and - high reliability. The superior service level is conducive to business activities in modern business districts. Chilled water supply from in-building plants is typically not metered. The metering and invoicing aspect of district cooling service heightens the awareness of the users on cooling energy usage, thus leading to greater conservation effort such as turning off air-conditioning when it is not needed. The Low Delta-T syndrome, if present in a building using the district cooling service, is reflected in the monthly invoice as the Return Temperature Adjustment for the Usage Charge. This surcharge also helps to alert the building management of the need for maintenance attention to the downstream facilities. Proper commissioning and maintenance of the downstream facilities will avert excessive flow in the downstream chilled water system, thereby removing the Low Delta-T syndrome and cutting down wasteful pumping power. VIII. CONCLUDING REMARKS A district cooling system creates economic value through its larger scale of operation. A quick growth in cooling demand to build a critical mass is essential for its commercial success. While the justification of its introduction to low load density areas, such as residential districts or a sprawling area with low-rise buildings, would be challenging in terms of economic viability, it is evidently a desirable utility service for high load density areas in an urban environment, particularly in the tropics. The strategic decision of the authorities in Singapore in fostering the successful implementation of the district cooling system at Marina Bay has yielded tangible outcome in elevating the energy efficiency for airconditioning of the new business district with greater certainty and effectiveness. VII. DEMAND SIDE ENERGY EFFICIENCY District cooling, with its utility attributes, is an enabling infrastructure for better energy efficiency at the utilization level. It helps to avoid these wasteful practices commonly observed in buildings: - Proliferation of less energy-efficient split unit airconditioners for use at off-peak period when the in-building chiller plant is not operating. - Operating additional AHUs than necessary at offpeak period to provide a minimal load to the chiller for stable operation. - Need for warm clothing in a tropical environment. Often some part of the building is over-cooled when centralized AHUs serve large floor area. The round-the-clock availability of chilled water from district cooling facilitates the adoption of localized control through the use of smaller AHUs or fancoil units. (Updated: June 2010)