HRSGs at Keppel I power plant in Singapore
HRSGs at Keppel I power plant in Singapore
Credit: Alstom

The changing role of combined-cycle gas turbine installations is prompting boiler designers to come up with a new generation of flexible steam generators capable of fast startup and rapid rampdown. These developments are finally setting free the innate flexibility of the gas turbine, writes David Appleyard

Even by normal standards, the far-reaching transformation of the electricity sector has been dramatic, as the rise of renewables has wrought havoc among the established conventions of the power generation system.

Indeed it is the capacity for change – and rapid change at that – which is singling out those technologies which are expected to do well in the new energy landscape. Among them are a new generation of heat recovery steam generators (HRSGs).

It is well known that among the principal challenges of accommodating large volumes of variable-output renewables is the need for a flexible generation network, which can respond to significant fluctuations in available capacity. It is also no secret that government and regulatory authorities are now struggling to develop effective market mechanisms that can support such a system and the consequent requirements for investment. There are also a number of major engineering challenges which must be addressed.

The innate flexibility of the gas-fired turbine, capable of ramping output up or down in a matter of minutes, lends itself to such a role. With the exception of geographically limited hydropower, gas turbines are the most flexible source of conventional power generation, but there are nonetheless limitations even here. For instance, where gas turbines are allied with a HRSG in a combined-cycle configuration, a number of issues can arise that significantly constrain the capability of the turbine.

These challenges include thermal inertia and, more significantly, thermal fatigue and increased wear in key components. For instance, the typical boiler design features at its heart a high-pressure steam drum. With its thick walls, such a vessel is designed to withstand high temperatures and pressures, but rapid changes in output can result in a number of effects chiefly related to the appearance of temperature differentials.

For example, admitting large volumes of cooler water into the steam drum as the plant ramps down and steam requirements decrease can see a temperature differential arise which, as a result of the thermal expansion properties of the steel, can result in the phenomenon known as ‘drum hump’ – a visible bend across the length of the high pressure vessel. Inevitably this places components under considerable additional strain with through-wall stresses potentially too high, resulting in reduced fatigue life of the boiler and premature failure.

And this effect can be found all over a steam generator with its rafts of tube bundles and their joints and bends. It means that in standard boiler designs, operators have to limit the rate of change in the gas turbine’s output in order to keep within the design limits for thermal gradients. This presents a number of difficulties for plant owners and operators. When gas-turbines are running at less than optimal output, as well as losing efficiency some emissions – particularly NOx – can increase significantly. Ideally, plant operators want to be able to ramp up their turbines as quickly as possible, not only to maximize market opportunities but also to benefit from the cleanest and most efficient operational mode achievable.

Naturally the designers and manufacturers of HRSGs have responded, with Dutch company NEM, for example, taking an early lead with its Drumplus design.

Gerard Van Dijk, chief executive officer at NEM, explains that the ‘DrumPlus’ design is an HRSG that allows a power plant to start up from cold in 10 minutes. “It allows the plant to respond very fast to fluctuations in the grid due to the electricity production of renewables,” he says. “A [coal-fired] power plant will not be able to respond to that quick enough; as a matter of fact, most combined-cycle power plants will not be quick enough to respond.” Van Dijk argues that a startup time of a matter of minutes “is essential for the future also of renewable energy”.

One way the company achieves this is through the design of a vertical once-through boiler which, rather than containing a high-pressure steam drum, is instead drumless. This enables structural integrity to be maintained under the same steam conditions, but using thinner material for the enclosure. Thinner material reduces the fatigue impact from varying thermal expansion and suffers less from through wall stress.

Commissioned a little more than a year ago, El Segundo, near Los Angeles in California, marked the first installation of a DrumPlus HRSG. Owned by NRG Energy, El Segundo produces a peak power of 550 MW and delivers more than half of its generating capacity in less than ten minutes, thereafter reaching full output in less than an hour. The El Segundo project saw the retirement of a 335 MW steam boiler unit constructed in 1964. The new unit uses some 30 per cent less natural gas per unit of electricity produced than the original boilers.

Other manufacturers are pursuing drumless designs. John DiVitto, business development manager for HRSGs at US company Babcock & Wilcox, explains: “We’ve had to make design changes to be able to be flexible and take into account the different physical changes that are occurring in the HRSG due to these rapid ramping events, both up and down.”

His colleague Larry Hiner, manager of boiler product lines for B&W’s global power division, outlines the major features of its natural circulation design: “We’ve made very flexible tube bundles. We have a lot of care that we put into the attemperator design to be able to handle the change in loads. Same with the drain systems for this type of unit: it’s not just a maintenance type drain, but is integral to its ability to start up very quickly.”

The design, known as a fast convertible vertical separator, features a pair of vertical HP steam drums of around half the previous required diameter.

Hiner says: “Hanging this vessel at the side of the unit allows the expansion and contraction to be in line with the tube bundles so it simply expands and contracts with the tube bundles and you’re relieving all the stresses on down-comer and connections.

“We have been utilizing this type of technology on UP or super critical boilers for some time, and more recently we’ve utilized this technology on a concentrated solar technology plant that we have installed out in California where they see very high cycling. The B&W drumless HRSG design was launched around two years ago and has been actively marketed for about a year.”

In another design approach, Jean-Francois Galopin, chief technology officer at Belgian HRSG manufacturing group CMI, explains that they too are currently developing a patent pending system to limit the drum wall thickness. Galopin adds: “We are improving our calculation capabilities to better simulate the startups. The first steps of the pressure increase are the most important, but most difficult to exactly simulate.”

Like B&W, CMI is actively exploring the potential for high-cycling HRSGs in solar thermal applications. Galopin says: “We have recently created a business unit dedicated to solar energy, we are finalizing the installation of a 50 MW solar boiler in South Africa and we’re starting the design of a 110 MW molten salt solar boiler in Chile.”

El Segundo in California has a peak power of 550 MW
El Segundo in California has a peak power of 550 MW
Credit: NEM

Minimal lifetime impact

For its new units, Alstom has developed a design that enables cyclic capabilities with minimal lifetime impact. However, it has adopted a different approach to the challenges of rapid ramping. Erwing Calleros Micheland, product manager for Alstom Power, says: “What we have seen with our analysis of the HRSG is that the most difficult component in cyclic operation is not the drum, but the HPSH and RHTR outlet manifolds. They are thick-walled components that are subject to the highest temperatures at full load operation and to the most rapid rates of temperature change during startup. So by modifying the design of those specific components, we have taken care of any issues of cycling the HRSG.”

He adds that Alstom has been using this design approach for around two decades.

Micheland outlines the key features of the so-called OCC design (optimized for cylcing and constructability), specifically the area between the tubes and the headers: “We have re-designed these components to reduce by as much as half the stress that is caused by temperature difference between the tube and the header compared to a conventional design.

“In addition, with the OCC we are using single row harps, tubes with no bends and thin-walled headers. This stepped component thickness change allows us to reduce the stress from differential thermal growth between tubes, headers and manifolds. Since we don’t have bends there are no bending forces which cause stress in the HRSG.”

A second key area of development is the increasing use of sensors and monitoring to more effectively manage the HRSG over its lifetime. Pascal Decoussemaeker, product manager for HRSG service at Alstom in Switzerland, says the company places considerable emphasis on advanced monitoring systems that help mitigate remaining risks related to cyclic operation and also enable the quantification of the cost of cycling a HRSG unit.

Decoussemaeker cites a typical example found in the startup of a warm or hot unit. Under such conditions, condensate is formed. If this cooler condensate builds up in the lower headers or manifolds, it can result in damaging top to bottom temperature differentials. Using multiple skin thermocouple sensors to monitor these temperature excursion events can allow them to be curtailed by modifying the operational profile, improving the drain system or allowing operators to more effectively manage the maintenance requirements of the key components that are affected.

A number of new products have emerged that support these functions. For example, NEM has launched its Wave radar – a drum level measurement by wave radar with no risk of frost (conventional systems freeze in heavy winters) and far greater accuracy.

CMI has also developed new management tools. Two years ago it launched Boiler Stress Evaluator and Galopin says: “This software allows a cumulative damage to be evaluated based on EN12952 formulations and various boiler inputs, such as metal temperature and pressure measurements, located at critical positions, such as high pressure final superheater components [thick and subject to creep] or high pressure drums [thick]. It has been offered on two of our last projects and is available for installation on any existing boiler that needs a fatigue evaluation due, for example, to a modification of its use.”

Decoussemaeker also flags up the role of a service and maintenance strategy when changing the operational profile of an HRSG. “On the service side we are confronted with units that are 10 or 20 years old and that have been designed for a more steady mode of operation.”

He argues that operators should make a thorough review of their HRSG from the moment flexible operations commence: “You look at your unit and you try to retrofit it, not by changing the complete design, but by addressing the areas where you expect the main problems – in the lower headers, lower manifolds and drain systems.

“Often by optimizing or redesigning the drain system you can avoid condensate flooding. Another problem that can occur is caused by the increased need to desuperheat. Over spray from desuperheater water injection could potentially damage downstream superheater tubes. If you know this can happen and if you redesign your desuperheater system, this can be avoided. These are smaller redesigns that help to avoid problems with the reliability of the unit.”

Further down the line, as Alstom’s Calleros explains, “We see a trend for bigger gas turbines. State-of-the-art gas turbines are bigger, they have higher exhaust flows, higher temperatures, so we see the need for larger HRSGs. We see that HRSGs are producing more steam, of course, at higher pressures and higher temperatures, and this is mostly driven by the need to increase the combined-cycle efficiency.”

Van Dijk echoes the point, saying: “The next step is to go for higher efficiencies on the power plant. That means power and efficiency of the gas turbine, but also higher power output and the efficiency of the HRSG steam turbine combination. In that case the only direction is to go for higher temperatures and pressures. That means we’re moving on what is physically possible with the current available materials, moving towards higher alloy materials. That’s where I see the developments going – higher quality alloys, different welding technologies, higher pressures and stresses etc.”

CMI has already supplied two stainless steel platens (final reheater and superheater) for a boiler which is currently being erected, for example.

Market outlook

It is clear that overcapacity in Europe has subdued the market. However, Van Dijk says: “We do expect developments in the Middle East. It is still in need of power – they have the oil and dollars to finance their projects. We also see developments in the US, where gas is available and there is a modest recovery of the economy, as well as in South East Asia and China.”

Similarly, B&W’s Hiner says: “Globally we see a strong potential for growth in gas, subject to the availability at various locations. Gas plants are typically cheaper and quicker to fabricate and start up, so that gives them a faster market strategy for a lot of these developing countries to bring on capacity equipment. Looking at between 2018 and 2035, the forecast is for about 420 GW of natural gas additions, a large percent we hope will be combined-cycle plants. We also see international growth.”

Ravi Krishnan, founder of consultancy firm Krishnan & Associates, gives his outlook: “Obviously the sale of HSRGs is a function of how big the gas market is and what the price of gas is relative to other fuels. Basically, in the US market, the price of gas is at a point where it makes it very attractive for new builds, largely because gas prices have been hovering at $4.50/mBTU and $6 depending on the season. At these prices it still makes it quite competitive for a developer to invest in gas-fired generation. Additionally, environmental rules are not exactly favouring new coal-fired generation, at least in the US. The price of gas along with GDP growth is what will result in new gas-fired capacity coming online.”

He highlights narrowing capacity reserve margins in markets such as Texas, which along with the low gas price (prompted by the shale gas boom) is triggering new investment. Krishnan says that roughly 6000-7000 MW of new gas-fired capacity already comes online each year in the US, a figure which is expected to grow based on GDP growth. He also highlights the potential of LNG to drive new HRSG installations – picking out countries such as Australia, Japan, Malaysia and Thailand as potential markets set for significant gas-fired growth.

“Once the LNG terminals are up and running in the US – which are scheduled for 2015 and 2016 – you can anticipate there will be significant export of LNG, which would still be [cost] viable in certain markets such as India and other parts of Asia where the use of gas is going to become even more prevalent.

“The role of gas in global power generation is only going to be going up.”

David Appleyard is a freelance journalist focused on the energy and technology sectors.

More Power Engineering International Issue Articles
Power Engineering International Archives
View Power Generation Articles on PennEnergy.com