In October last year, the UN issued a landmark report which stated Greenhouse gas emissions must be cut almost in half by 2030 to avert global environmental catastrophe, including the total loss of every coral reef, the disappearance of Arctic ice and the destruction of island communities. Scientists stated that we need keep temperature increases below 1.5°C to avoid the worst effects of global warming. Unless we see a significant move away from fossil fuels to renewable energy and introduce new technology to reverse global warming by removing CO2 from the atmosphere, scientists believe this figure will be exceeded within around 20 years.
This report paints a bleak picture of our future if rapid action is not taken on an individual and corporate level. Organisations have a responsibility to shareholders to look after them both financially and from an environmental perspective so they need to start taking a long-term view and base purchasing decisions on reducing environmental impact. In this way, we can all play a part in contributing to change.
So, from an environmental perspective what needs to be considered when purchasing a UPS?
If we consider a 100kW UPS operating 24 hours per day/365 days per year, every 1% of efficiency loss wastes 1kW every hour. At only 10p/kW hour this equates to £8,760 over a 10 year period and does not include the energy wasted by additional air conditioning. Currently in the UK, more than 60% of this wasted energy comes from the burning of fossil fuels. This is why selecting the most efficient UPS is essential to minimise the carbon footprint.
Beware, operational efficiencies are often stated by manufacturers as being ‘greater than 99%’, however, this 99% efficiency is related to ‘offline’ operation or ‘ecomode’. No datacentre ever operates on this mode as they would be operating on raw mains and only transferring (with a short break in power) to full UPS operation when there is a problem. True online efficiency is the important percentage to compare UPS solutions as this indicates the real UPS operating efficiency.
From an efficiency perspective, the big challenge is that the IT power requirement in most organizations will change over time. Six-years ago a Comms Room server rack typically required 6kW, today it is likely to require 30kW. So how can the infrastructure be built to meet these dramatically changing demands?
Put simply, a UPS needs to be easily scalable and must always operate at the optimum point on its efficiency curve. A system which is too small will be overloaded, compromising availability while a system which is too large will be inefficient, waste energy and be costly to run. It will also cost more than necessary to maintain due to its size. Scalability and flexibility are therefore essential considerations when purchasing, to ensure the continual ‘right sizing’ of the UPS.
State of the art UPS technology, such as CENTIEL’s 4th Generation modular technology CumulusPower, has a flat efficiency curve for loads above 15% thereby offering the combined benefits of increased flexibility, scalability and lowest running costs.
Most UPS systems can operate in ambient temperature of 40°C without de-rating, however, VLRA (lead-acid) batteries used in UPS start to degrade at above 20°C. For every 10°C above 20°C, the useful working life of the battery is halved. So, a VRLA battery with a ten-year design life working at 20°C will last five years at 30°C and just over two years at 40°C. In addition to the battery replacement costs and/or the requirement for air conditioning there is the environmental cost of replacing many tonnes of batteries that current legislation classifies as ‘special waste’.
An alternative is Li-ion battery technology. Li-ion has a higher purchase price but because Li-ion batteries can operate at higher ambient temperatures without degrading the need for air-conditioning is very significantly reduced. In Northern European locations such as the UK this means UPS cooling could be provided by the natural air temperature resulting in significant savings on data centre running costs and, equally importantly, reducing its carbon footprint.
Li-ion batteries also have a much longer useful working life. Five-year design life VRLA batteries operated at 20°C are normally replaced every 3-4 years. With Li-ion this is 13-15 years.
When it comes to UPS purchase there are usually two budgets: one for the capital expenditure and one for maintenance. Traditionally, the employee selecting the cheapest UPS were misguidedly praised for ‘saving money’ but the cheapest UPS are invariably built with the cheapest components which have much higher repair and maintenance costs and are therefore actually ‘costing money’. A top-quality UPS such as CENTEIL’s CumulusPower, using Li-ion batteries will need only one capacitor change 15 years whereas an inferior solution will typically need three capacitor changes plus three sets of replacement VLRA batteries in 15 years.
Commercial UPS installations can learn from Industrial UPS installations where >25 years’ working life is normal for a UPS. It is true that IT technology is changing very rapidly but the same 230V / 50 Hz being used 20 years ago will be the same in 20 years.
So, by using UPS that are engineered to last, commercial organisations can contribute positively to environmental impact by reducing the need for maintenance, replacement parts and scrappage.
Correct Decision Making
The increase in UPS efficiency from 85% in the 1980s to 97% today is a 12% improvement that reduces the UPS’ carbon footprint but purchase decision makers are often still focused only on the bottom line, missing the bigger environmental picture. Of course, increased efficiency and lower total cost of ownership are closely linked and the most environmentally friendly systems enjoy ongoing operating cost savings. However, given a choice, decision makers still take an ill-informed short term view and purchase the lowest cost system thinking they are “saving money” for their company and often contrary to the company’s environmental and sustainability policies. While this behaviour is understandable it is “old school” and needs to change from the top down and employees need to be incentivized and educated help the make the right environmental choices for all our sakes.
For further information please visit www.centiel.co.uk
EkkoSense, the data centre M&E Capacity Planning and Simulation specialist, has launched a major partner recruitment drive to help accelerate its expansion plans – both in the UK and internationally.
EkkoSense has identified the five data centre optimisation trends that it believes will make a real difference for organisations in 2019.
Hydro66, founded in 2014 is a pioneering, ultra-efficient, green-field colocation data center located in Boden, Northern Sweden. A key objective of the company was to design their new Nordic data centre to operate at a power usage effectiveness (PUE) – the ratio of total amount of energy used by a data centre to the energy delivered to computing equipment – of less than 1.05. This could only be achieved using fresh-air cooling (EcoCooling ECT10800 Nordic Cooling range) supported by the most efficient uninterruptible power supply (UPS) and power distribution.
Hydro66 decided to use a direct ventilation system supplemented by evaporative cooling (EcoCooling ECT10800 Nordic Cooling range). The equipment is modular and installed internally, thereby avoiding planning issues. Electronically commutated (EC) axial fans are used for air movement. With very low pressures, axials can also accommodate the larger flow rates and pressure, and their motors are efficient, quiet and have simple speed controls. The efficiency of a fan is approximately proportional to the cube of the speed. Data centres require redundancy of N+1, 2N or 2(N+1), so equipment is operated at part capacity.
By controlling all of the EC fans as a group – and reducing the air flow rate to that required by the IT equipment reductions in consumer fan power can be achieved, producing remarkable efficiencies. On average, 1MW of IT equipment will require an airflow of 90m³/s of air at compliant temperatures. Since the data centre has both redundancy and spare capacity, the ventilation rate is reduced and further savings are made. For example, running a fan at 80% reduces energy use by half and, at 50%, to 12.5%. An intelligent control system is used by Hydro66 constantly to optimise the fan energy use to reflect actual cooling requirements in a dynamic environment. On warmer days, the adiabatic cooling is enabled, bringing the supply air down to approach the wet-bulb temperature of the ambient air. In Boden, this means the supply air will never exceed 22C, which is compliant with all standards without the need to use additional mechanical refrigeration.
Hydro66 had a very clear vision on how we could bring a new model to colocation – one where the customer wins significantly on both cost and on sustainability. We were fortunate to discover EcoCooling who were able to exceed our expectations. Not only in terms of pure efficiency of their equipment, but more importantly their desire and capability to enhance their solutions to our specific use case.
ALEX CHIOLO, HYDRO66 OPERATIONS DIRECTOR
The use of adiabatic cooling will increase the moisture content, while reducing dry-bulb temperature, so increasing the relative humidity of the air. With reference to the ASHRAE 2011 Thermal Guidelines, high relative humidity (RH) will normally only cause corrosion with other contaminants in the air. If gases such as sulphur or chlorine are in the ambient air, these, plus high RH, can cause corrosion. Boden has ‘clean’ air because there are no local industries producing contaminants.
The combination of high RH and dust or particulates can also create problems, so all incoming and recirculating air is filtered. In relatively clean conditions such as those in Boden, EU4 is a suitable level of filtration. Increasing this can result in significant increases in capital cost, maintenance requirements and fan energy use. A direct fresh-air system
A direct fresh-air system operating in arctic conditions at the coldest time of the year can result in very low RH in the data centre. Low RH, in conjunction with other factors, can cause problems with electrostatic discharge (ESD), which can damage IT equipment. EcoCooling specially designed Nordic cooling system incorporates a recirculation loop, where – in low RH conditions – the warm air from the data centre is passed over the adiabatic pads to humidify the air above the ASHRAE 2011 Thermal Guidelines’ allowable level of 20%. This novel solution, therefore, uses the adiabatic pads for two functions – cooling in hot weather and humidification in cold weather conditions.
Hydro66 has constructed a low capital cost, flexible data centre, which has achieved a PUE of less than 1.05. The direct fresh-air Nordic cooling system complements the Download the data sheet for the ECT10800 internal evaporative cooler with humidification. More information on data centre cooling
Etix Blockchain, a division of Etix Group dedicated to provide colocation services for HPC and Blockchain applications and mining as a service capacity, specified a large number of EcoCooling ‘3 CloudCooler Group’ modules for their new facility in Iceland. EcoCooling were selected based on the success of their CloudCooler Range in other installations across the region, which has become a hotbed for data centre development mainly due to competitive power prices and cool climate.
With its ambient air conditions Iceland is an ideal location for data centers using the EcoCooling solution. A number of award winning data centre and cryptocurrency mining operators in the Nordics and Arctic circle have used EcoCooling solutions to achieve unparalleled cooling efficiency.
“Opening two large-scale Blockchain data centers in Iceland in a very short period of time, we had to call on a trustworthy partner with solid experience to support us in this challenge. We have been cooperating successfully with EcoCooling to provide our customers highly-efficient data centers.”
Antoine Gaury, Head of Etix Blockchain
“We are very excited to be providing the cooling equipment for Etix Blockchain in their new facility in Iceland. Our new CloudCooler Group (rack, cooling and containment) solution is a true plug and play mining infrastructure module. The simple design significantly increases speed of installation so operators like Etix Blockchain can scale up their facilities in an extremely cost effective and energy efficient way.”
- Alan Beresford, MD, EcoCooling
The ‘3 CloudCooler Group’ module has been developed off the back of these installations to be the most cost-effective option for rapid deployment developments in remote locations. The design complements Etix Blockchain’s scalable data centre design strategy while also maintaining world class cooling efficiency.
Each group installed contains 3 ECV CloudCoolers®, containment, power supply and racking. The ECV coolers provide 54,000m3hr of air directly to the miners while the Group design was chosen for its simplicity and scalability. The full solution including containment took just 2hrs to assemble from flat pack.
Estimated performance: The average energy use to support 5.2MW of computing load is estimated to be approximately 173kW as an annual average with a peak of 340kW. This represents a pPUE of 0.033 for the cooling system.
ECV CloudCoolers - Free Cooling Units
ECV CloudCoolers (free cooling) units can maintain a reliable operating environment for computing devices in Iceland due to the cold climate. The ambient temperature rarely exceeds the supply temperature required by the computers. This means filtered, external air can be used all year round to cool the facility, eliminating the need for a supplementary cooling system.
EcoCooling has extensive experience in terms of cooling blockchain systems, cryptocurrency miners as well as conventional IT equipment both in Arctic and temperate climates. Using fresh air ventilation systems, with or without adiabatic cooling, compliant conditions are maintained for ~40kW cooling energy per 1MW of installed miners. This minimises both operating costs as well as the cost of the power infrastructure to support the cooling system. www.ecocooling.co.uk
About Etix Blockchain
Etix Blockchain offers colocation services for HPC and Blockchain applications, and Mining as a Service capacity through a network of data centers located in the Nordics. Etix Blockchain benefits directly from the expertise developed by Etix Everywhere for the data centre industry and from the security products developed by Etix Labs. www.etixblockchain.com
About the CloudCooler Range
With the ability to withstand cold climates, the UK manufactured CloudCooler® products come fully equipped with exiting and innovative technologies, proving a perfect match for the many emerging and established European data centres looking to capitalise on the green energy available in the Nordics.
The CloudCooler® units provide a constant supply of filtered air at controlled temperatures. Filtration plus the avoidance of temperature fluctuations maximise the reliability of the computers and other IT equipment. Utilising these key areas provide a dependable platform to maintain the maximum availability of mining power.
EkkoSense (D1210), the UK-based data centre thermal software and services expert, has launched the world’s first Internet-of-Things (IoT) enabled wireless thermal sensor. The new ultra-low cost sensor disrupts the traditional data centre sensor cost model, enabling it to be deployed in such vast numbers that true real-time thermal management of critical facilities such as data centres now becomes realistic.
EkkoSense - stand D1210 at Data Centre World 2018 - is to launch the world’s first Internet-of-Things (IoT) enabled wireless sensor to provide a local display of measured temperature and relative humidity values.
The EkkoSensor Wireless Display Sensor can show temperature profiles over the last hour, 24 hours or week for immediate on-site thermal assessment, and integrates with EkkoSoft Critical 3D software to provide real-time virtual reality visualisation.
Etix Labs, a new global data center innovation provider, today launches the most comprehensive suite of Data Center Infrastructure Monitoring (DCIM) tools in the industry, combining software, hardware, and services to provide data centers managers intelligent infrastructure monitoring, physical security and video surveillance management. Etix DCIM, Keynesia and Etix Video Management System provide a brand new experience to track everything and anything happening in mission critical infrastructures.
Proactive managed service contract signed for 5 of Daisy’s key UK data centre sites
– reduced risk, increased capacity, 19% energy savings
UK-based data centre thermal risk expert EkkoSense has released EkkoSoft Critical 3.2 – the latest version of its SaaS-enabled 3D thermal visualisation and monitoring software offering for managing critical data centre thermal risks.
Excel Networking Solutions is excited to announce that they will be hosting a series of Consultant Briefings across the country in February 2018.
The EU research and innovation programme, Horizon 2020, has awarded a three-year innovation project in Northern Sweden. The aim is to prototype the most energy and cost-efficient datacenter in the world.
EcoCooling has walked away with a high commendation at the 13th annual RAC Cooling Industry Awards held in London.
Energy saving ultrasonic humidifiers installed at Manchester data centre
The DataSafe XE range is suitable for high temperature environments American energy storage specialist EnerSys has announced a new range of uninterruptible power supply (UPS) batteries, designed to de ...
What do data centres require of their power delivery? The answer is easy: 100% power availability, 100% of the time. In other words, they want a system that never fails.
However, reliability is often considered the key attribute for any UPS solution. Yet, a system can be reliable over a period of time, but still fail on a particular occasion with far-reaching consequences. Therefore, availability must be the number one priority when purchasing any UPS solution.
Availability can be expressed by the following equation where MTBF equals Mean Time Before Failure and MTTR equals Mean Time To Repair:
Availability = MTBF
MTBF + MTTR
Availability is expressed as a percentage. The key number here is MTTR (if MTTR = 0 then Availability is always 100% regardless of MTBF).
In recent years, modular systems have introduced a significant step-change in the industry as, when properly configured, they are designed to maximise load availability and system efficiency simultaneously.
This is achieved because modular systems have a single frame, containing a number (N) of power modules. These run together and share the load equally between them. The advantage of this N+1 configuration is that if one module fails it becomes isolated. The remaining modules support the load and the system availability is preserved. UPS modules are paralleled vertically within a single frame, and frames can be further paralleled horizontally to provide a completely flexible system. There is no single point of failure, contributing to the highest level of availability for power protection. The load also remains protected even when any individual module is being replaced.
CENTIEL’s CumulusPower is the latest 4th generation of true modular UPS systems, incorporating Distributed Active Redundant Architecture (DARA), it provides a significant improvement over previous system designs. The critical power protection industry is, understandably, extremely riskaverse, and is particularly attentive regarding potential single points of failure. Therefore, each individual CumulusPower UPS module contains all the elements of a standalone UPS unit – rectifier, inverter, static switch, display – and critically – all control and monitoring circuitry. This eliminates the potential single point of failure associated with designs that have a separate, single static switch assembly, separate control or intelligence modules. Availability for the 4th generation Centiel system is now Nine 9s or 99.9999999%.
In addition, with all the focus on the UPS units themselves it can be forgotten that an integral element of a system is the associated batteries which provide power in the event of an electrical supply interruption or blackout. Equal consideration needs to be given to the design of the battery system, including rigorous battery autonomy calculations, selecting a quality supplier, the provision of redundant strings, individual string protection/isolation, ambient environment to name but a few.
With all the time, effort and expense of designing and installing the very best UPS system it is important that it is properly maintained to ensure availability is not compromised. Suppliers offer a variety of maintenance proposals and SLAs and diligence must be applied up front, in scrutinising and choosing the correct solution for each installations requirement. Any UPS that is not maintained properly may ultimately compromise overall system availability.
As well as availability, organizations need to consider the total cost of ownership of a UPS system and this is related to the purchase price, system efficiency (running costs) and the cost of maintenance.
I’ve touched on maintenance, so how a system is configured to maximize efficiency is another area demanding attention. The challenge is that IT requirements change quickly. So how can an infrastructure be built to meet these dramatically changing demands? Put simply, a UPS needs to operate on the best point of its efficiency curve. A system which is too small will be overloaded and a system which is too large will waste energy and be inefficient and costly to run. Scalability and flexibility are key when purchasing, to ensure the continual ‘right sizing’ of the UPS.
Modular systems offer all the advantages of high availability, scalability and flexibility to match an organization’s changing requirements. In addition, they offer the benefits of: fast replacement of modules, small footprint with high power density and, when correctly configured to ensure system “right sizing,” operating costs can be kept to a minimum.
In our ever-evolving world, future-proofing systems is one of the greatest challenges faced by system designers. The good news is that modern modular UPS systems can be quickly and easily reconfigured to adapt to changes in load requirements over time. This not only ensures the highest efficiency is maintained, but more importantly it guarantees availability of power protection whatever the future holds.
Article written by Mike Elms, Sales & Marketing Director, CENTIEL UK Ltd
Tier 3 and 4 data centres are designed with redundant power from source to socket, to cater for any eventuality resulting from a single failure. But power outages in data centres can still occur. Re ...