Tag Archives: convection cooling

Industry Developments: Cooling Electronics in Wind Turbines

By Norman Quesnel, Senior Member of Marketing Staff
Advanced Thermal Solutions, Inc.

(This article will be featured in an upcoming issue of Qpedia Thermal e-Magazine, an online publication dedicated to the thermal management of electronics. To get the current issue or to look through the archives, visit http://www.qats.com/Qpedia-Thermal-eMagazine. To read the preceding post on Cooling Solar Power Inverters, click https://www.qats.com/cms/2016/11/21/industry-developments-cooling-solar-power-inverters.)

Wind power systems capture natural air currents and convert them, first to mechanical energy and then electricity. Windmills have long harnessed natural, renewable wind currents to grind grains and pump water. Now those windmills have evolved into highly engineered wind turbines, with very long, highly-engineered blades spinning on steel towers some that are tens of meters high.

There are some relatively small wind turbines that power individual houses or businesses. They can generate around 100 kW of power. But most of today’s wind turbine industry is for utility-scale power generation. These are large, tall wind turbines, in fields of dozens or hundreds, delivering high levels of electricity to power grid systems that reach thousands of end users. More than a quarter million of such turbines are in use around the world.

Cooling Electronics in Wind Turbines

Fig. 1. The Alta Wind Energy Center in California has more than 600 wind turbines and can produce more than 1.5 GW of power. [1]

Most utility-scale wind turbines are built on open, naturally windy land or off-shore. Each turbine can produce 1.0-1.5 MW, enough energy to power hundreds of homes. The United States has about 75 GW of installed wind power capacity. And, despite some local resistance, the U.S. has begun to join other countries with off-shore installations. China has by far the most installed wind power capacity at about 150 GW. Globally, the combined power capacity from wind turbines is forecast to nearly double between 2016 and 2020 to 792 GW. This would be enough to power 220 million average homes in the U.S. [2, 3]

Mechanics of Wind Turbines

When natural wind blows past a turbine, its blades capture the energy and rotate. This rotation spins a shaft inside the rotor. The shaft is connected to a gearbox that can increase the speed of rotation. The gearbox connects to a generator that produces electricity. Most wind turbines consist of a steel tubular tower. On top of this is a nacelle structure, housing the turbine’s shaft, gearbox, generator and controls.

On the wind-facing end of the nacelle is a hub to which the turbine blades are attached. Together, the blades and the hub are called the rotor. The diameter of the rotor determines how much energy a turbine can generate. The larger the rotor, the more kinetic energy is harnessed. Furthermore, a larger rotor requires a taller tower, which exposes the rotor to faster winds. [4]

A wind turbine is equipped with wind assessment equipment, including weather vanes. These send data to a computer to automatically rotate the turbines into the face of the wind and to a pitch system that can angle the blades to further optimize energy capture. [5]

Cooling Electronics in Wind Turbines

Fig. 2. The major components of a wind turbine. [6]

Turbines and Fire

Hundreds of wind turbines catch fire each year. The most common cause is lightning strikes, but overheated equipment can also be responsible. Highly flammable materials such as hydraulic lubrication oil and plastics are in close proximity to machinery and electrical wires inside the nacelle. A fire can ignite from faulty wiring or overheating. The results are catastrophic. The rush of oxygen from high winds can quickly expand a fire inside a nacelle. Once a fire starts, it is not likely to be deliberately extinguished. Water hoses can’t reach a nacelle’s height and wind turbines like these are typically set in remote locations, far from emergency aid. [7]

Cooling Electronics in Wind Turbines

Fig. 3. A wind turbine’s blazing nacelle and hub at a wind farm in Germany. Lubricating oil is often the fuel when these fires occur. [8]

Electronic Devices in the Nacelle – and Heat

Most wind turbines don’t catch fire, of course. Yet, despite all the surrounding wind, the electronics in their nacelles still need significant thermal management to function continuously. The most important electronics are the generator and power converting devices.

The generator is the heart of a wind turbine. It converts the rotational energy of the wind-spun rotor into electrical energy. It generates the electric power that the wind turbine system feeds into the grid.

Generating electricity always entails the loss of heat, causing the generator’s copper windings to get hot. Larger capacity generators are even further challenged. The thermal losses will increase with the generator in proportion to the cube of its linear dimensions, resulting in a serious decline in generator efficiency.[9]

Excess generator heat must be dissipated to maintain efficiency and avoid damage. On most wind turbines this is accomplished by enclosing the generator in a duct, using a large fan for air cooling. Some manufacturers provide water-cooled generators that can be used in wind turbines. The water-cooled models require a radiator in the nacelle to void the heat from the liquid cooling matrix.

Wind turbines may be designed with either synchronous or asynchronous generators, and with various forms of direct or indirect connection to the power grid. Direct grid connection means that the generator is connected to the (usually 3-phase) alternating current grid.

Wind turbines with indirect grid connections typically use power converters. These can be AC-AC converters (sometimes called AC/DC-AC converters). They change the AC to direct current (DC) with a rectifier and then back to usable AC using an inverter. In this process, the current passes through a series of Insulated Gate Bipolar Transistor switches (IGBTs). These convert direct current into alternating current to supply to the grid by generating an artificial sine wave. The more frequently the switch is turned on and off, the closer to a true sine wave the current flow becomes, and the more sine-like the flow, the purer the power. The resulting AC is matched to the frequency and phase of the grid. [10]

However, the faster these switches actuate, the more heat they develop and given a wind turbine’s variable inputs, IGBTs for this application need to cycle very frequently. This generates large amounts of heat that will dramatically decrease overall efficiency unless properly cooled. [11]

Cooling Electronics in Wind Turbines

Fig. 4. An active air cooling system inside a wind turbine nacelle features an air-to-air heat exchanger for managing heat in the generator (Vensys). [12]

Even with efficiency improvements, a wind turbine’s power generation systems and subsystems must manage ever increasing heat within its limited nacelle space. In addition, even if incurred power losses are as little as 3-5 percent, thermal management systems would have to dissipate 200-300 kW and more of heat.

Air cooling has been used effectively in small-scale wind turbines, but it is not practical for removing the heat produced in MW-scale units. Its thermal capacity is so low that it is difficult to blow enough air across a motor or through the converter to maintain reliable operating temperatures. That is why water cooling is used more often than air for larger wind turbines.

Cooling Electronics in Wind Turbines

Fig. 5. Electronics in a medium voltage (Up to 12 MW) wind turbine converter. Cooling is provided by a closed-loop unit with a mix of deionized water and glycol (ABB).[13]

However, water cooled systems are relatively large, and their thermal efficiency limitations force the size and weight of power generation sub-systems to essentially track their power throughput. Due to the thermal performance limitations of water, the power-generation equipment for a 10 MW wind turbine is nearly twice the size and weight of a 5 MW model. This is largely because water cooling cannot adequately remove additional heat loads without spreading them out.

One supplier of liquid cooling systems for wind turbine electronics is Parker Hannifin. Its Vaporizable Dielectric Fluid (VDF) system provides heat transfer capability significantly greater than that of water. The VDF system requires less fluid and lower pump rates. The same dissipation rates provided by a 6 liter/minute water flow can be achieved by 1 liter/minute VDF flow, thus allowing for a smaller system.

The hermetically sealed VDF assembly is designed to be leak proof, but if a leak occurs the non-conductive fluid will not damage electronic components. The cooling system’s efficiencies and lack of thermal stack-up provide an additional advantage in that the system maintains a fairly tight temperature range. The lack of thermal cycling removes a strain on the turbine’s electronics, which extends their useful life. [14]

Cooling Electronics in Wind Turbines

Figure 6. Dual-phase liquid cooling method for converters has a circulating refrigerant in a closed-loop. Vaporizing coolant removes heat from devices and re-condenses to liquid in a heat exchange (Parker). [15]


Heat issues in wind turbine electronics mainly concern the generator and the power conversion electronics. The heat load of the generator comes from copper wire resistance and from iron loss from the rotation of the core. Further heat loss is mechanical due to friction. These energy losses become heat energy that is distributed into the wind turbine nacelle.

The excess heat from the nacelle-based power conversion systems is mainly due to impedance from electronic components such as capacitors and thyristors. Higher temperatures will reduce the system’s life and increase failure rate. Thermal management methods such as liquid cooling can be effectively adapted for nacelle electronics. [10]

1. https://en.wikipedia.org/wiki/Alta_Wind_Energy_Center
2. https://en.wikipedia.org/wiki/Wind_power_by_country
3. http://www.ozy.com/fast-forward/how-twinning-tech-will-power-our-future/71993
4. Layton, Julia, How Wind Power Works, HowStuffWorks.com.
5. http://www.awea.org/Resources/Content.aspx?ItemNumber=900&navItemNumber=587
6. https://www.linkedin.com/pulse/smart-grid-energy-harvesting-martin-ma-mba-med-gdm-scpm-pmp
7. http://www3.imperial.ac.uk/newsandeventspggrp/imperialcollege/newssummary/news_17-7-2014-8-56-10
8. https://www.youtube.com/watch?v=sYoQ6mS2gss
9. http://ele.aut.ac.ir/~wind/en/tour/wtrb/electric.htm
10. Jian, S., Xiaoqian, M., Shuying, C. and Huijing, G., Review of the Cooling Technology for High-power Wind Turbines, 5th Intl Conf on Advanced Design and Manufacturing Engineering, 2015.
11. http://www.windpowerengineering.com/design/mechanical/cooling-electronics-in-a-hot-nacelle/
12. http://www.vensys.de/energy-en/technologie/generatorkuehlung.php
13. https://library.e.abb.com/public/430f5f2493334e4ead2a56817512d78e/PCS6000%20Rev%20B_EN_lowres.pdf
14. http://www.windsystemsmag.com/article/detail/60/cool-system-hot-results
15. http://buyersguide.renewableenergyworld.com/parker-hannifin-renewable-energy-solutions/pressrelease/parker-to-launch-converter-cooling-systems-for-1mw-wind-turbines-at-husum-wind-energy-2012.html

For more information about Advanced Thermal Solutions, Inc., its products, or its thermal management consulting and design services, visit www.qats.com or contact ATS at 781.769.2800 or ats-hq@qats.com.

Case Study: LED Solution for Outdoor Canopy Array

Advanced Thermal Solutions, Inc. (ATS) was approached by a company interested in a new design for an outdoor LED unit that would be installed in gas station canopies. The original unit was bolted together and contained a molded plastic shroud that held the LED array, the PCB, and an extruded aluminum heat sink.

ATS engineers designed an aesthetically pleasing alternative that utilized natural convection cooling, while increasing the number of the LEDs in the array and its power. The engineers met the customer’s budget and thermal performance requirements.

Challenge: Create an outdoor canopy device that would increase the number of LED in the array, increase power to maximum of 120 watts, and increase lumens, while cooling the device through natural convection.

Chip/Component: The device had to hold an LED array and the PCB that powered it.

Analysis: Analytical modeling and CFD simulations determined the optimal fin efficiency to allow air through the device and across the heat sink, the spreading resistance. The weight of the device was also considered, as it would be outside above customers.

Solution: An aesthetically-pleasing, one-piece, casted unit with built-in electronics box for LED array and PCB was created. There was one inch of headroom between the heat sink and the canopy to allow for heat dissipation and the casting would allow heat transfer as well as allow air to flow through the system.

Net Result: The customer was able to add LEDs to the array and increase power. The new unit also simplified the manufacturing process and cut manufacturing costs.

For more information about Advanced Thermal Solutions, Inc. thermal management consulting and design services, visit www.qats.com or contact ATS at 781.769.2800 or ats-hq@qats.com.

Case Study: Thermal Management in Harvard Medical School Tissue Analysis Instrumentation

Designers of today’s highly advanced medical diagnostic equipment must overcome many of the same thermal challenges common to telecommunications, industrial and information technology electronics.

In addition, medical diagnostic devices present unique design issues and boundary conditions that factor into thermal solutions. These include isothermal and cyclic temperature demands, precise test repeatability, and maintaining the patient’s safety and comfort.

These kinds of issues were presented by Harvard Medical School to the experts at Advanced Thermal Solutions, Inc. (ATS) when it needed a cooling solution for the diagnostic equipment it was relying on for the analysis and observation of human tissue samples in a controlled laboratory setting. This was the school’s Frozen Tissue Microarrayer System.

ATS engineers had to provide thermal solutions to meet a range of design goals:

• Provide long-term temperature control for tissue samples embedded in an optimum cutting temperature fluid.
• Create a cooling system to maintain tissue samples below -70°C for six hours.
• Ensure operator visibility of the samples.
• Eliminate humidity and frost within the system to prevent sample contamination.

ATS Cooling Solutions

ATS engineers developed highly effective thermal solutions to meet all the design requirements of the diagnostic equipment. A reservoir in the device holds the liquid cooling medium and tissue samples are loaded through an opening at the top. Through a duct, cool air is circulated over the top of the samples to maintain temperature and humidity requirements.

As seen in Fig. 1 (below), the diagnostic system consisted of:

• Frozen tissue coring machine (on the right in the photo)
• Tissue sample loading area at the top of the cooling system (seen on the left)
• Duct system (on both sides of system) to circulate cool air
• Ice/alcohol reservoir at the system’s bottom to contain the cooling medium

Harvard Case Study

Figure 1. Prototype system created by ATS engineers for Harvard Medical School laboratory. (Advanced Thermal Solutions, Inc.)

Conduction Cooling Design

In operation, tissue samples are loaded into removable aluminum cassettes that fit tightly into a metal receiver (top left, Figure 2). The receiver contacts the cassette on five sides which allows for cooling of the samples by conduction. The receiver is lowered into a reservoir containing a slurry of dry ice and ethyl alcohol. Here the receiver is maintained at a constant temperature until the dry ice evaporates. The reservoir is double-walled and insulated to extend the evaporation time of the dry ice.

The receiver also features integral fins that increase surface area for drawing heat downward from the base of the cassettes into the icy slurry (bottom left, Figure 2). These fins are based on the same ATS heat sink design principles used in the company’s high performance maxiFLOW™ heat sinks.

Using analytical modeling, ATS engineers determined that 10 fins were the optimal number for cooling the cassette receiver and its contents. CFD simulations also showed that the 10-fin concept resulted in an optimal design. The engineers validated their analytical and CFD results through empirical testing. It was determined that extending 10 fins into the slurry provided the cooling performance to maintain tissue sample temperatures below the -70°C threshold for 9.75 hours.

Further temperature testing using thermocouples showed only a 2.5°C difference between the coldest points at the bottom of the fins and the tissue samples in the cassette. This proved that the design overcame thermal conduction resistance and could effectively maintain the samples below their critical temperature.

Figure 2. Temperature testing with thermocouples demonstrated that the temperature difference between the bottom of the fins and the top of the cassette, through three intervening layers, was only 2.5°C. This proved that the thermal design was successful. (ATS)

Figure 2. Temperature testing with thermocouples demonstrated that the temperature difference between the bottom of the fins and the top of the cassette, through three intervening layers, was only 2.5°C. This proved that the thermal design was successful. (ATS)

Figure 3. Using a heat sink-specific thermal resistance network ATS determined that the optimal number of fins was 10. (ATS)

Figure 3. Using a heat sink-specific thermal resistance network ATS determined that the optimal number of fins was 10. (ATS)

Convection Cooling Design

The above conduction cooling design provided only part of the solution. There were additional needs to maintain the temperature at the top of the samples and to decrease the relative humidity of the cool air from the ambient air in the lab. ATS engineers designed a convection cooling system to fulfill these requirements.

A heat exchanger was installed with its fins in the dry ice/alcohol slurry and its other side extending into a duct to cool the air passing over it. This approach uses the same cooling medium for both convection and conduction to ensure there is no temperature differential throughout the sample and that the sample is as isothermal as possible.

Air is pushed by a counter-rotating fan through the duct and into the heat exchanger. The heat exchanger forms a thermal link between this air and the slurry mixture. The heat exchanger was designed with an optimum balance between its surface area and the resulting pressure drop to ensure the fan was operating with the most effectiveness.

Once the air passes the heat exchanger, it moves through the ducts and into a diffuser at the top of the system. The diffuser disperses the air over the sample creating a barrier between the tissue and the ambient environment of the lab so outside moisture and heat are not transferred in.

The ATS engineers tested this design using an array of thermocouples and ATS hotwire anemometer Candlestick Sensors connected to an ATS ATVS-2020, a temperature and air velocity scanner. They determined there was too much mixing between the air flowing over the samples and ambient air. The diffuser was redesigned with a new connection to the duct and an optimized outlet radius (see Figure 4).

In the ducts, a molecular sieve desiccant housed in a honeycomb structure was used to reduce the dew point of the air to -84.4°C, which was well below the -72°C air temperature in the duct.

Figure 4. Initial testing led to a redesign of the air diffuser to prevent ambient humidity from mixing with the air over the tissue samples. (ATS)

Figure 4. Initial testing led to a redesign of the air diffuser to prevent ambient humidity from mixing with the air over the tissue samples. (ATS)


ATS engineers performed a final series of tests of the Frozen Tissue Microarrayer System using Candlestick Sensors, thermocouples and the ATVS 2020 scanner. The tissue temperature stayed constant over the required six-hour period and well below the -70°C threshold. In fact, testing determined that the tissue temperature remained below the threshold for nearly eight hours before warming above a usable temperature (Figure 5). The multi-part cooling system was a success, meeting the original design objectives provided by Harvard Medical School.

Figure 5. Final testing showed that the ATS cooling design kept tissue temperature (shown in blue in the graph above) below the -70°C threshold for more than the required six hours. (ATS)

Figure 5. Final testing showed that the ATS cooling design kept tissue temperature (shown in blue in the graph above) below the -70°C threshold for more than the required six hours. (ATS)

The process of designing cooling solutions for the Frozen Tissue Microarrayer demonstrated that thermal design practices used throughout electronics cooling can be applied in the medical device industry. Fin efficiency, thermal resistance, and pressure drop calculations are standard regardless of the application. Thermal solutions should be considered early in the design process so they can be incorporated into the overall system as efficiently as possible.

The experts at ATS used traditional thermal calculations, CFD simulations, empirical testing, and its leading-edge heat sink technology to successfully design the thermal solution for this medical equipment application. The ATS design allowed Harvard Medical School to test tissue samples while meeting its strict requirements.

To learn more about the design, watch the video below:

Download a PDF of this case study at http://www.qats.com/cms/wp-content/uploads/Harvard-Medical-case-study.pdf.

Visit www.qats.com, call 781-769-2800 or email ats-hq@qats.com to learn more about ATS and its Thermal Management Analysis and Design Services.

The Importance of Heat Flux Sensors

Heat flux sensors are practical measurement tools which are useful for determining the amount of thermal energy passed through a specific area per unit of time. Measuring heat flux can be useful, for example, in determining the amount of heat passed through a wall or through a human body, or the amount of transferred solar or laser radiant energy to a given area.

Affixing a thin heat flux sensor to the top of a component will yield two separate values which are useful in determining the convection heat transfer coefficient. If the heat flux can be measured from the top of the component to the ambient airstream and if the temperature at the top of the component and of the ambient airstream is measured, then the convection coefficient can be calculated.


q = Heat flux, or transferred heat per unit area

h = Convection coefficient

TS = Temperature at the surface of the solid/fluid boundary

TA= Ambient airstream temperature

Using a heat flux sensor can be useful for lower powered systems under natural convection scenarios. Under forced convection, the heat lost to convection off the top of a component can often be significantly higher than the heat lost to the board, particularly if the board is densely populated and the temperature of the board reaches close to the temperature of the device. Under natural convection situations, often the balance of heat lost to convection and heat lost through the board becomes more even and it therefore is of even greater interest to the designer to understand the quantity of heat dispersed through convection.

Experiments done at Bell Labs alluded to the effect of board density on the heat transfer coefficient. In these experiments, thin film heat flux sensors are affixed to DIP devices which populate a board. The total heat generation of the board is kept constant, so the removal of components from a densely populated board only increases the heat generation per component. The results of this particular experiment highlight an increase in the ratio of heat lost through convection from the surface of the component as board density increases and individual device power decreases.

 Surface Heat Flow vs. Board Density

Qs/Qt = the ratio of total heat flow through device surface to total heat generation

σ = the ratio of total device surface area to total board area

If a board was to be sparsely populated, a greater percentage of heat can be transferred to the board due to larger thermal gradients; however since the overall surface area of the sum of devices decreases, to some extent the heat transfer coefficient must increase to reflect a balance. As the number of components decreases, the power generation increases per component, and the larger resulting temperature gradients in the region around the component yield more convective flow and thus an increase in the heat transfer coefficient. On the other hand, if the board becomes more densely populated, the proportion of heat transferred through the surface, as compared to through the board increases, and the overall increase in heat transferred through the surface yields increased flow and heat transfer at an individual component surface.

The use of a heat flux gauge is an important tool for the electronics designer. In particular, by using a heat flux gauge, it is possible to experimentally determine the heat transfer coefficient at a certain location on the electronics board where it would have had to be simply predicted or estimated previously. Due to the complexity of many electrical systems as well as the irregular nature of many boards, often analytical or CFD methods are not accurate and the best approach is empirical techniques. The use of the heat flux sensor can give results which would be difficult to calculate using analytical or numerical simulations. However, like most other instruments, it is important to use the sensor correctly and carefully to decrease the errors within a system and increase the reliability.

How to Apply Natural Convection Cooling for the Thermal Management of Electronics (part 3 of 3)

In part 1 of our 3 part series, we covered why and how natural convection works and how a heat sources geometry can affect natural convection.  Part 2, addressed board orientation and its impact on natural convection cooling and useful equations that can be used for calculating initial and simplistic results based on viable assumptions. In part 3, our final part in this series, we‘ll cover the calculations in generic plate fin heat sinks and wrap up our series.

For a generic plate fin heat sink, with a fin height of H, and spacing between fins of B, there is a relative fin spacing to fin length ratio which will provide optimal heat transfer within a given volume. Where spatial constraints prevent the use of a larger sized heat sink, optimizing its design for natural convection and increasing its efficiency can be the difference between normal operation and thermal failure.

straight fin aluminum heatsink

Figure 3. A Simple Plate Fin Heat Sink.

Because the heat sink has to include heat transfer from the base, an equivalent channel width needs to be determined. The equivalent channel width is solved similar to a hydraulic diameter, as seen below [2].

The equivalent channel width math formula


H = fin height

B = distance between fin surfaces (fin pitch minus fin thickness)

The heat extracted from the heat sink can be found using this equivalent channel width and an assumed temperature difference between the ambient and heat sink surface.

.  Heat Dissipation of a Sink for Different Configurations of Fin Spacing and Fin Height Natural Convection

Figure 4. Heat Dissipation of a Sink for Different Configurations of Fin Spacing and Fin Height Natural Convection [2]

The results shown in Figure 4 highlight the effects of changing the fin height relative to fin to fin spacing. For low volumes, it may not be economical to extrude or machine a heat sink with the exact geometry to optimize heat transfer from a given control volume, but the results show that for a standard or off-the-shelf solution it is safer to pursue a design with a larger fin spacing to fin height ratio than otherwise.

In this article we have shown that many applications prevent the use of forced convective flow from fans or other sources, due to acoustics, cost or other reasons. But, optimizing heat transfer by using the natural buoyancy-induced movement of air can be effective in cooling electronics. Some techniques for optimizing airflow and heat transfer in natural convection situations involve a vertical channel. For heat-generating boards suspended in a vertical configuration, there is calculable board spacing which will lead to an optimal power density per volume. Heat sinks can also be optimized for natural convection, and while it may not be feasible to obtain a heat sink of the perfect fin height and fin pitch for a workable volume, the next best performer is a heat sink with larger fin spacing relative to fin height.


1. Hwang, P., Cheng, H., Fang, J. and Li, J., CFD-Based Thermal Characterization of Board-Level Microelectronic Devices under Natural Convection Cooling, Microsystems, Packaging, Assembly and Circuits Technology, 2007. IMPACT 2007.

2. Malhammar, A., Optimum Sized Air Channels for Natural Convection Cooling, Telecommunications Energy Conference, 1987.

3. Incropera, F., Liquid Cooling of Electronic Devices by Single Phase Convection. New York, Wiley. 1999.