Energy Recovery Inc: pressure exchanger technology?

pressure exchanger

A pressure exchanger transfers energy from a high-pressure fluid stream to a low-pressure fluid stream, and can save up to 60% input energy. Energy Recovery Inc is a leading provider of pressure exchangers, especially for the desalination industry, and increasingly for refrigeration, air conditioners, heat pump and industrial applications. Our technology review finds a strong patent library and moat around Energy Recovery’s pressure exchange technology.


Energy Recovery Inc was founded in 1992, it is headquartered in California, listed on NASDAQ, with 250 employees and $1.3bn of market cap at the time of writing. Financial performance in 2022 yielded $126M revenues, 70% gross margin, 20% operating margin.

The PX Pressure Exchanger is Energy Recovery Inc’s core product. It transfers pressure energy from a high pressure fluid stream to a low pressure fluid stream at 98% efficiency, yielding up to 60% energy savings in specific contexts. The company aims to grow revenues as much as 5x in the next half-decade due to increasing need for global energy efficiency.

Pascal’s Law states that bringing a high pressure and low pressure fluid into contact will result in their pressures equalizing with minimal mixing. This principle is used in pressure exchangers. As a rotor rotates, it brings a low pressure fluid A into contact with a high pressure fluid B, equilibrating their pressure, then discharging fluid A at higher pressure.

For example in a desalination plant, incoming seawater at 1-3 bar of pressure is pressurized up to 40-80 bar using pumps, pushing it across a membrane that is porous to water but not to dissolved salts. Energy remains in this 40-80 bar concentrate stream. It is better to recover this energy than blast it back into the Ocean! Thus pressure exchange can lower the energy requirements of desalination by as much as 60%.

Energy Recovery Inc’s patents note that rotary pressure exchangers were first invented in the 1960s, progressed in the 1990s, but prior to its own designs, the company argues that no one had designed efficient and reliable systems, which could run without an external motor to rotate them, achieved by optimizing the shape of the flow channels.

Our technology review found 65 patent families from Energy Recovery Inc. Overall, we think the patent library is high-quality and the company will retain a moat and leadership in the pressure exchange market, based on its patents and historical experience. Although some early patents are coming up to expiry. Details are in the data-file.

Desalination has been Energy Recovery’s core market historically. However new markets are emerging, from cryogenic cycles through to applications focused on shale (although the latter requires avoiding the corrosive impacts of sand and debris in fluid streams).

pressure exchanger
End markets for pressure recovery based on patents filed by Energy Recovery Inc.

Refrigeration, air conditioning and heat pumps are seen as a growing source of demand. One patent notes that regulation is increasingly phasing out HFCs that can have 13,000x higher GWPs than CO2, and these systems use CO2 as the refrigerant instead. However CO2 based refrigeration cycles have maximum pressures of 1,500 psi or greater, compared to HFC/CFC systems at 200-300psi. This makes the energy savings from pressure exchange increasingly important, siphoning away a portion of the evaporated refrigerant and re-pressurizing it using high-pressure refrigerant downstream of the heat rejection stage, before the expansion valve stage.

LEDs: seeing the light?

Outlook for LEDs

Lighting is 2% of global energy, 6% of electricity, 25% of buildings’ energy. LEDs are 2-20x more efficient than alternatives. Hence this 16-page report is our outlook for LEDs in the energy transition. We think LED market share doubles to c100% in the 2030s, to save energy, especially in solar-heavy grids. But demand is also rising due to ‘rebound effects’ and use in digital devices. We have screened 20 mature and (mostly) profitable pure plays.

LED lighting: leading companies in LEDs?

Leading companies in LED lighting

20 leading companies in LED lighting are compared in this data-file, mostly mid-caps with $2-10bn market cap and $1-8bn of lighting revenues, listed in the US, Europe, Japan, Taiwan. In 2022, operating margins averaged 8%, due to high competition, fragmentation and inorganic activity. The value chain ranges from LED semi-conductor dyes to service providers installing increasingly efficient lighting systems as part of the energy transition.


The global LED industry is worth $80bn per year, with LEDs used in lighting indoor and outdoor spaces, in the automotive industry and to back-light the screens associated with the rise of the internet.

Leading companies in LED lighting range from specialists manufacturing semi-conductor dyes, other specialists combining these components with other conductors and phosphors into LED packages, others encasing this product into LED lamps, others adding further drivers and housings to yield LED luminaires, and others ultimately selling entire lighting systems.

Integration versus fragmentation. Some companies are involved in the entire, integrated value chain discussed above, while others specialise in specific parts, e.g., dye/phosphor specialists upstream, or enterprise solutions businesses that design and implement overall lighting systems for corporate customers.

This data-file profiles 20 leading companies in LED lighting and the broader lighting industry. For each company, we have quantified size, recent financial metrics (e.g., revenue, operating margins), estimated exposure to the LED lighting industry and tabulated key notes. Backup tabs in the file cover LED costs, LED payback periods and LED efficiency.

Competition is high in the LED lighting industry and the average company reported an 8% operating margin in 2022.

Fragmentation is high, as the largest company in the screen derives just $7bn per annum of LED-related revenues. All 20 companies in our screen generated c$40bn of revenue in 2022.

Company sizing is therefore more skewed towards mid-cap and smaller-cap names than other company screens we have undertaken, with the larger companies in the screen having market caps in the range of $2-10bn.

Acquisition activity in the LED lighting industry has been high, and half a dozen of the companies in our screen have recently changed hands or been spun out. Details are in the data-file. For example, Philips Lighting was IPO’ed as Signify in 2016 and remains the largest integrated LED lighting company in the world.

Industry leaders in LED lighting include listed companies in the Netherlands, Japan, the US, Taiwan, China, Austria and Germany.

AirJoule: Metal Organic Framework HVAC breakthrough?

Montana Technologies is developing AirJoule, an HVAC technology that uses metal organic frameworks, to lower the energy costs of air conditioning by 50-75%. The company is going public via SPAC and targeting first revenues in 2024. Our AirJoule technology review finds strong rationale for next-gen sorbents in cooling, good details in Montana’s patents, and challenges that decision-makers may wish to explore further.


The global HVAC market is worth $355bn per year, as air conditioning consumes 2,000 TWH per year, or 7% of all global electricity (note here).

Each one of the world’s 1.7bn vehicles also has an in-built air conditioner, and in hot/congested conditions, the AC can sap 70% of the range of an electric vehicle.

Typical air conditioners have a coefficient of performance (COP) of 2.5-3.5x, which means that each kWh of electricity inputs will deliver 2.5-3.5 kWh-th of cooling energy.

Cooling energy, in turn, is used to reject heat from air (about one-third of the total) and to reject heat from the inevitable condensation of water, as air cools down (about two-third of the total). To re-iterate: (a) cooler air can hold less water; (b) hence some water will condense as air is cooled; (c) condensation of water releases energy (called the latent heat of vaporization, and running at 41 kJ/mol, or 640 Wh/liter); and (d) this adds to the cooling load that the air conditioner needs to provide; (e) by a factor of 1-2x more than the cooling load needed to cool down the air in the first place (called the specific heat of air, or 1.0 kJ/kg-K). For more details, please see our overview of air conditioning energy demand.

Metal Organic Frameworks are a type of sorbent with a high surface area, able to adsorb water from air. To re-release that water, a vacuum pump can be used to lower the pressure (chart below). Effectively this is a swing adsorption process rather than a cryogenic process.

How does AirJoule technology work? This data-file pieces together details from Montana’s disclosures and patent filings. In particular, the work covers five core patents describing a Latent Energy Harvesting System, using Metal Organic Frameworks to adsorb and re-vaporize atmospheric water; test stable MOFs, and novel methods for manufacturing those MOFs.

What COP for AirJoule? After reflecting the loads of the vacuum pump, other fans and blowers, heat transfer from an adsorption chamber to a desorbtion chamber, and auxiliary loads, AirJoule is targeting a total electricity use of 60-90 Wh/liter of water. Hence if water is adsorbed, then desorbed, then misted, it will absorb 640Wh/liter as it re-vaporizes (see above). Divide 640Wh/liter by 60-90Wh/liter, and you derive a coefficient of performance of 7-10x. Or in other words, the energy consumption will be 50-75% below a typical air conditioner. More details are in the data-file.

What are the key challenges for AirJoule technology? We see five thematic challenges for using MOFs and sorbents in cooling systems, and two specific challenges based on reviewing Montana’s patents.

Overall, we think there is very strong potential for next-gen sorbents, and metal organic frameworks, across cooling and other industrial processes, as described in our recent research note here. Key conclusions into Montana’s technology are in our AirJoule technology review below.

Omniscience: how will AI reshape the energy transition?

AI reshape energy transition

AI will be a game-changer for global energy efficiency. It will likely save 10x more energy than it consumes directly, closing ‘thermodynamic gaps’ where 80-90% of all primary energy is wasted today. Leading corporations will harness AI to lower their costs and accelerate decarbonization. This 19-page note explores the opportunities.

EROEI: energy return on energy invested?

Energy return on energy invested

Net EROEI is the best metric for comparing end-to-end energy efficiencies, explored in this 13-page report. Wind and solar currently have EROEIs that are lower and ‘slower’ than today’s global energy mix; stoking upside to energy demand and capex. But future wind and solar EROEIs could improve 2-6x. This will be the make-or-break factor determining the ultimate share of renewables?

Energy efficiency: a riddle, in a mystery, in an enigma?

Projections of future global energy demand depend on energy efficiency gains, which are hoped to step up from <1% per year since 1970, to above 3% per year to 2050. But there is a problem. Energy efficiency is vague. And hard to measure. This 17-page explores three different definitions. We are worried that global energy demand will surprise to the upside as efficiency gains disappoint optimistic forecasts.

Prime movers: efficiency of power generation over time?

Efficiency of power generation over time

How has the efficiency of prime movers increased across industrial history? This data-file profiles the continued progress in the efficiency of power generation over time, from 1650 to 2050e. As a rule of thumb, the energy system has shifted to become ever more efficient over the past 400-years.


In the early industrial revolution, mechanical efficiency ranged from 0.5-2% at coal-fired steam engines of the 18th and 19th centuries, most famously Newcomen’s 3.75kW steam engine of 1712. This is pretty woeful by today’s standards. Yet it was enough to change the world.

Electrical efficiency started at 2% in the first coal-fired power stations built from 1882, starting at London’s Holborn Viaduct and Manhattan’s Pearl Street Station, rising to around 10% by the 1900s, to 30-50% at modern coal-fired power generation using pulverised, then critical, super-critical and ultra-critical steam.

The first functioning gas turbines were constructed in the 1930s, but suffered from high back work ratios and were not as efficient as coal-fired power generation of the time. Gas turbines are inherently more efficient than steam cycles. But realizing the potential took improvements in materials and manufacturing. And the best recuperated Brayton cycles now surpass 60% efficiency in world-leading combined cycle gas turbines.

Renewables, such as wind and solar, offer another step-change upwards in efficiency, and will harness over 80% of the theoretically recoverable energy in diffuse sunlight and blowing in the wind (i.e., relative to the Betz Limit and Shockley-Queisser limit, respectively).

There is a paradox about many energy transition technologies. Long-term battery storage and green hydrogen would depart quite markedly from the historical trend of ever-rising energy efficiency in power cycles. Likewise, there are energy penalties for CCS.

The data-file profiles the efficiency of power generation over time, noting 15 different technologies, their year of introduction, typical size (kW), mechanical efficiency (%), equivalent electrical efficiency (%) and useful notes about how they worked and why they matter.

Gas turbines: operating parameters?

Gas turbine operating parameters

A typical simple-cycle gas turbine is sized at 200MW, and achieves 38% efficiency, as super-heated gases at 1,250ºC temperature and 100-bar pressure expand to drive a turbine. The exhaust gas is still at about 600ºC. In a combined cycle gas plant, this heat can be used to produce steam that drives an additional turbine adding 100MW of power and c20% of efficiency, for a total efficiency of 58%. This data-file tabulates the operating parameters of gas turbines.


Why do gas turbines matter? Recuperated Brayton cycles are going to be a defining technology of the energy transition and a complement to renewables. The thermodynamics are explained here. The key point is that gas-fired power cycles are totally different from steam cycles. They run off a fuel that is 50% lower carbon than coal. They can realistically be 2-3x more efficient per unit of fuel. They are more flexible (data here). And they may also be easier to decarbonize directly (example here).

How does a gas turbine work? First, air is drawn into a compressor. The compression ratio is typically around 20x. The pressurized air is then heated by combusting a fuel. The result is a very hot, very high-pressure gas. This can be used to drive a turbine as it expands. For example, expanding 1 ton of gas from a turbine inlet temperature of 1,250ºC and a turbine inlet pressure of 100-bar, down to an exhaust gas temperature of 600ºC and near-ambient pressures, might see volumes increase by around 25x (chart below).

Simple cycles versus combined cycles. If the 600ºC exhaust gas is simply discharged into the atmosphere, then a typical simple cycle gas turbine will achieve 38% efficiency, converting natural gas into electricity. But there is still a lot of energy in a 600ºC exhaust stream, which can be used to evaporate water, produce high pressure steam, and then drive an entirely separate turbine. This is a combined cycle configuration. And it adds another 20% efficiency, yielding a total efficiency of 58%.

Note that the steam cycle described above, powered by the waste heat from a gas turbine, is effectively the same as the primary heat cycle used in other conventional thermal power plants (Rankine cycle). This is remarkable.

The efficiency of a simple cycle gas turbine depends primarily on the turbine inlet temperature and pressure, which in turn depend on the compression ratio. The most efficient simple cycle gas turbines hit 43% efficiency, with compression ratios of 25-30x, turbine inlet pressures of 140-180 bar and turbine inlet temperatures of 1,400-1,600ºC. It is quite hard to get hotter than this, because things start to melt. But consider, for contrast, that a steam cycle really struggles to surpass 300-500ºC.

Why does a gas turbine look like that? To achieve these high compression ratios a typical gas turbine will have 12-22 separately optimized and sequential compression stages. And to maximize power output in the turbine, it will typically have 4 turbine stages. This explains the classic cross sectional profile of a gas turbine.

How fast does a gas turbine spin? A simple cycle gas turbine typically spins at 3,000-4,000 revolutions per minute (rpm). The compressor is connected to the same shaft as the turbine. The back-work ratio imparted to the compressor is equivalent to around 40-50% of the net work driven through the turbine.

How large is a gas turbine? A typical 200MW gas turbine might take up 60 m2 and weigh 300 tons. Good rules of thumb are 0.3 m2/MW of areal footprint, and 2 tons/MW of weight. Although larger gas turbines are more compact (on a per MW basis).

What is the cost of a gas turbine? A typical gas turbine might cost $200/kWe (chart below). Larger gas turbines have lower costs per MW (chart below). However note that our model of a gas-fired power plant assumes total capex of $850/kW. In other words, total installed capital costs are typically around 4x larger than the turbine itself.

(This multiple may be worth keeping in mind amidst debate about hydrogen electrolyser costs. Some companies have been guiding to $200-300/kWe electrolyser selling prices, and some analysts noting that this realistically means around $1,000-1,200/kW fully installed costs).

Emissions from natural gas power plants are generally low. CO2 intensity is 0.3 kg/kWh from a 60% efficient combined cycle gas turbine (up to 70% below coal power plants). NOx emissions are usually below 25ppm but can be as low as 2ppm in the best models. Many new turbines are also hydrogen ready, and have been qualified for 25-75% hydrogen blending.

Flexibility of a gas fired power plant is middling to high. A typical plant can ramp up or down by 15% of its nameplate capacity per minute, turn down to c25-50% of its load, and start up from cold in 20-minutes. Different examples are tabulated in the data-file.

Our outlook for gas turbines in the energy transition is published here. Leading companies in gas turbines are profiled here. Gas turbine operating parameters are compiled for a dozen gas turbine models in this data-file, as a useful reference, mainly designs from Siemens Energy, GE, Mitsubishi-Hitachi and Ansaldo.

Thermodynamics: Carnot, Rankine, Brayton & beyond?

Thermodynamic cycles

Engines convert heat into work. They are governed by thermodynamics. This note is not a 1,000 page textbook. The goal is to explain different heat engines, simply, in 13-pages, covering what we think decision makers in the energy transition should know. The theory underpins the appeal of electrification, ultra-efficient gas turbines, CHPs, nuclear HTGRs and new super-critical CO2 power cycles.

Copyright: Thunder Said Energy, 2019-2023.