Batteries, Hydroelectric

Canyon Hydro Raises the Bar in Hydro Impeller Machining

Issue 2 and Volume 116.

By Chris Murray, field applications specialist, Ingersoll Cutting Tools

A CNC machining center makes a Deming, Wash., company one of the country’s leading supplier of large hydropower generating systems west of the Mississippi – right in the heart of the hydropower market. Advanced tooling and heavy customer support from Ingersoll Cutting Tools was instrumental in this accomplishment.

As a result of this partnership, Canyon Hydro completes huge ten-ton rotors (called runners in the hydropower industry) in one-third the time as before, and to much better accuracy and surface finish for higher efficiency over the runners’ projected 30-40 year service life. The runner is the heart of a hydropower turbine, converting water flow to the rotary motion that drives the generators.

Buckets Require Long-Reach Contouring

The key machining challenge is to contour-mill the runner buckets, essentially the paddles on a very sophisticated paddle wheel where every machined surface has a hydrodynamically correct curvature. A typical runner, made from a stainless steel casting, weighs 10 tons, measures 11 feet across and has 22 carefully contoured double buckets. It can take months of continuous five-axis, long-reach milling to complete, leaving behind a ton and half of stainless steel chips. Virtually all milling involves 19-inch shank lengths. Nevertheless, at the end, as-machined surfaces must be smoother than 32µ and geometrically correct within 0.010 inch, in order to prevent turbulent flow while in service.

“It’s like a large-scale cavity milling job, with one important difference,” said Ingersoll’s Chris Murray, who devised the tooling solution. “The contour on each bucket involves undercuts, which you would never see in an injection mold.”

Strategic Process Change

Previously, Canyon Hydro, headquartered in Deming, Wash., finished the buckets by manual grinding, using CAD/CAM templates to check dimensions and contours. Typical cycle times for an 11-foot runner were six months, involving 100 percent attendance by skilled operators. Typical tolerances were 0.035-0.040 inches. The process was more than adequate for its time. It was the era of low oil costs and easier environmental regulations which left hydropower as a secondary source of power along with solar, wind and nuclear.

Anticipating greater demand for alternate energy sources including hydropower, the company made the strategic decision in 2009 to automate runner machining. They built a free-standing, lights-out capable CNC center just a few towns away in Sumas.

The key machine is a FPT five-axis CNC Floor Type Horizontal Mill equipped with a two-axis rotary table, 80-station automatic tool changer and two interchangeable heads: a 360° Universal 3 + 2 Bi-rotational head and a 29 inch extension head. A skeleton crew of two to four CNC machinists works eight-hour shifts to handle support functions as the machine runs largely unattended, sometimes overnight as needed to maintain delivery schedules. “Chipmaking itself is essentially hands-off,” said manufacturing engineer Mike Hansen.

Accuracy, Surface Quality Trump Cycle Time Savings

Hansen invited tooling solution proposals from all mainline vendors, stressing the long-reach aspect along with the need for lights-out process security, exceptional accuracy and standard tooling. “Cycle time savings were really secondary to repeatable accuracy and smooth finish,” he said, “as we knew the business would become more efficiency-competitive.” Standard tooling was essential to eliminate all the uncertainties and expenses inherent in special tooling, Hansen said.

Nevertheless, most vendors proposed special tools or asked for up-front money for development. Murray offered full application support for free and found a standard tool able to do most of the job and a modified-standard to handle the rest.

The company started up by automating the smaller runners, diameters down to 4½ feet, gradually working up to the larger parts. The smaller runners were handled by a standard 1 inch Ingersoll FormMaster Pro specially suited for long-reach roughing and finishing. To defeat harmonic vibration, the three-flute tool features circular, serrated inserts in a “timed” array. Each insert is turned “five minutes” from the other, so its edges engage a different area of the cut and the whole toolpath is covered progressively with every full cutter revolution. Pips in the seat pocket mate with dimples on the backside of the insert to keep it in exact position.

The operation ran smoothly with no chatter, reducing cycle time on average by 50 percent versus manual grinding. Tool life was more than enough for lights out operation as needed.

Scaling Up Brings New Tooling Challenges

It was February 2011 before Hansen converted the larger runners over to the automated process and had to scale up the tool sizes accordingly. Principal model in this class is the 11-footer, with an average annual volume of 12 pieces.

Geometry of the runner buckets – six-axis contours long reaches most and some undercuts – necessitated a ball mill for the most part. Sheer size of the workpieces required a tool much larger than most standard models. “To do a three-foot contoured cavity with a one-inch ballmill or FormMaster would take forever,” Hansen said. “Besides, in stainless the insert would wear out too quickly for secure lights-out operation.”

Hansen recommended a completely standard 2-inch Ingersoll ProBall indexable ball nose for the bulk of the work and a modified standard Form Master button cutter to handle the undercuts. “The button cutter works like a standard contour mill for most of the pass, then like a T-slotter when it reaches the undercut portion along the outer edge of the bucket,” Murray said.

Setting Parameters

Hansen and Murray worked together, right at machine-side, to establish machining parameters for the 10-ton 410 NM stainless steel casting. With the big ballmill, the standard for mainstream roughing is 15 IPM, 650 RPM, 3/8 in. DOC, ¼ in. stepover. For finishing it is 45 IPM, 825 RPM, 0.050 in. DOC, 0.100 in. stepover.

The undercut portions required the modified button cutter and involved the longest reaches: 20.5 in. from spindle to face. Roughing parameters here are 30 IPM, 700 RPM, 0.375 in. DOC, 0.100 in. stepover. For finishing the settings are 45 IPM, 825 RPM, 0.050 in., DOC, 0.035 in. stepover.

Hansen used these parameters during the day when the site was manned, backing down the feedrate about 10 percent as a precaution for lights-out operations. In all cases, the inserts lasted long enough for absolute process security over 12 to 15 hours; some lasted 45 hours per edge.

The Key Tools, Close Up

The big two-flute ProBall ballmill features serrated inserts at the ball end plus heavy duty side cutting inserts farther up the active length for larger diameters and deeper cuts.

“Visitors to our shop floor are amazed that such a large ballmill comes standard,” Hansen said. “Without it we’d be faced with either a standard or some compromise tool.”

A screw-on style coupling enables in-spindle tip shuttling to 0.005 in. repeatability so there’s virtually no dead time for tool servicing.

More a modified standard than “special,” the button cutter is needed to access one portion of the cut which involves undercutting much like slotting. The backside of the tool must be effective upon withdrawal, to create the “top” of the slot. The tool uses standard inserts, with the cutter shank extended and insert seats repositioned to present cutting edges on both the front and backsides. Hansen gave Murray the CAD file on the starting and finishing geometry of the outer lip of the bucket, then Ingersoll product specialist Mark Teno and his team back at Rockford took it from there.

Ramping It Up – Carefully

Working together on the shop floor, Hansen, Murray and lead CNC machinist Ken Neal have bumped up the parameters about 15 percent overall since February 2011, always paying attention to accuracy, surface quality and process security in a very chatter-prone operation. “It’s as much a matter of listening as anything else,” said Murray. “I don’t think anybody can truly optimize a long-reach stainless steel milling operation over the phone.”

Not surprisingly, all of Canyon Hydro’s runners are completed much faster and with much less operator attention as a result of the company’s strategic move to CNC automation and advanced tooling. Cycle times average 50 percent less and labor costs are virtually nil. The mainstream 11 footer runs even better: 500 hours vs. 1,500 before. And the tool life is reliably long enough to enable lights out operation as needed.

Bridging the Gaps in Containment Systems

By Ed Sullivan, freelance technology writer

The modern design and construction of containment structures – both primary and secondary – are designed and constructed to be foolproof, ensuring safety to the workers, the public and environment. Yet, as safe as they may be, the eventual failure of containment system coatings and linings result in maintenance or rebuilding that is time-consuming and expensive, and may even require exorbitant downtime.

There are many choices of products that are designed to protect against leaking or spilling of the aqueous materials contained in sumps, trenches, concrete dikes and tanks with concrete bottoms. Yet many of the most frequently used products – polyurea, polyurethane, polysulfide, silicone and epoxy – offer only limited protection, and will fail due to exposure to UV light, weather, chemicals, abrasion and cracks in concrete or expansion and control joints.

Failures in primary containment structures, such as clarifiers, can result in extensive downtime. Many clarifiers are constructed of metal walls but have concrete bottoms, sometimes reinforced with grout to provide an extra seal if the concrete should crack from thermal expansion and contraction or other stresses. Yet, in the majority of cases, when the concrete moves or cracks, the grout cracks, causing leakage into the secondary enclosure and need for a cleanup. It also means the primary enclosure must be shut down and repaired.

While secondary containment dikes are required to enclose leaks and spills for three days, (72 hours), damage to the lining is likely from the spill of most chemicals resulting compromising the dike’s integrity, resulting in costly and time-consuming maintenance and downtime. UV rays and weather cause concrete to move and crack, also, and just plain degrade most polymers, causing the same problems.

There is a solution to the combined threats of UV, weather, chemical and expansion-contraction problems, an engineered elastomeric lining system that can be applied to primary and secondary containment structures ranging from wastewater facilities to protecting livestock ponds from chemical infiltration. The engineered elastomeric lining is noteworthy for its long service life, ability to “bridge” joints and cracks in concrete, imperviousness to UV light and harsh chemicals, and ease of installation.

The Elasti-Liner system from KCC is a line of engineered elastomeric lining products that is applied by brush or roller to concrete substrates and directly over expansion and control joints.

The elastomeric bridge

“In the past, we used 100 percent epoxy coatings on our secondary containment structures,” said Brian Peroni, Corrosion Control Specialist at Florida Power & Light. “Epoxy might work well enough for some primary containment applications, but for concrete structures, it’s inadequate. In concrete structures live cracks will occur, and because epoxies are a rigid coating, they tend to crack right along with the concrete. As a result, you have continual maintenance to seal those cracks as they occur.”

Peroni describes FP&L’s typical secondary containment as basically a concrete dike consisting of a wall and base around the facility’s primary tanks. Those tanks contain a variety of liquids, ranging from very harsh, low pH chemicals, such as 98 percent sulfuric acid, to very caustic solutions with a pH as high as 10 to 12. Tanks located in Florida are subjected to a lot of sunlight, which contributes to the cracking of concrete dikes.

The clarifiers at power plants are another example of containment structures where cracking can be a serious problem. These primary containment vessels are often designed with metal sidewalls resting on a concrete base. In an attempt to prevent leaks when the concrete base cracked, the concrete is often reinforced with grout. However, when the concrete moves or cracks, the grout seal also can fracture.

To overcome the problems of leaks resulting from cracked concrete in primary and secondary containment structures, Peroni uses and recommends Elasti-Liner elastomeric liner.

“The product has great crack bridging properties,” he said. “That’s why it is so effective for us. We don’t have to worry about cracks in the concrete and we don’t have a lot of maintenance. We simply put it on and forget about it.”

Peroni is an advocate of using suppliers who can provide the right materials for the application. KCC is one of the suppliers who he feels offers that support, which is why he feels that this is one of the leading suppliers in the industry.

The inside story

“I would say that this Elasti-Liner product line is the only truly monolithic containment liner available,” said Art Rak, president of Ultimate Corrosion Control.

Rak cites the 1994 installation of the elastomeric formula as a lining of a large, 4-ft. containment dike for an 18,000-sq.-ft. tank farm holding highly-corrosive phosphoric acid in Chicago Heights, Ill.

“The lining still looks and works great after 17 years,” he said.

Rak said this elastomeric product line is great at bridging cracks up to one-eigth inch because of the way it is engineered. The polymers are cross-linked and act like coiled springs that expand and also contract as concrete moves.

“The elastomeric liner exceeds the tensile strength of concrete,” he added. “If you deliberately tried to pull the liner off the concrete, you would have to exceed 500 PSI in pulling strength. So, this elastomeric-based lining is so well bonded to concrete no matter what the concrete does.”

Because you can apply Elasti-Liner over expansion and control joints in concrete structures, installation time and costs are saved.

“One customer told us that using the Elasti-Liner system, including the joint sealer, cost him less than it would have just to have an installer take care of sealing the joints using a standard method,” Rak explained.

About the author: Ed Sullivan is a Hermosa Beach, Calif.-based writer. He has researched and written about high technologies, healthcare, finance and real estate for over 25 years.

EPA’s Risk Management Plan Updated Guidance

By Michael Boozer, President, ChemReport, Inc

In 1996, the U.S. Environmental Protection Agency (EPA) promulgated the Risk Management Plan (RMP) regulations (40 CFR Part 68) under the authority of the 1990 Clean Air Act (CAA) to ensure the development and implementation of effective process safety management programs for facilities that store and use listed highly hazardous chemicals. EPA’s RMP regulations cover 430 power generating facilities; 354 of which are regulated due to the presence of ammonia (Fig. 1).

Power generation facilities typically use ammonia in Selective Catalytic Reduction (SCR) systems, as a refrigerant for cooling air intake in combustion turbines, and for water treatment systems. While this article addresses some of the nuances associated with aqueous ammonia use, the information presented applies to any chemical subject to RMP regulation.

EPA has developed various RMP guidance documents with respect to ammonia and other chemicals for certain industrial sectors. However, no specific industry guidance exists for the power generation industry. Consequently, power generators have been left with the task of interpreting various available RMP guidance documents for other industries with regard to the two chemicals in most prevalent use by SCR systems, anhydrous ammonia and aqueous ammonia solution.

EPA has issued comprehensive guidances on conducting RMP inspections and has updated many of the various guidance documents developed for other industrial sectors. Being cognizant of the new EPA guidance will go a long way toward successfully passing EPA inspections which are being conducted more frequently now that the RMP program has had 15 years to mature. Subjects of particular interest to the power generation industry include how to report the total process weight of aqueous ammonia solutions, the proper use and application of land classification systems, appropriate air dispersion modeling programs for release scenarios and trigger events that require the RMP to be updated.

EPA Inspections

In January of 2011, EPA’s Office of Enforcement and Compliance issued an 84-page guidance document for “Conducting Risk Management Program Inspections.” This guidance contains various checklists for inspectors to use for assessing a facility’s compliance status. It is an excellent resource for facilities to reference when conducting their own internal RMP audits, which are required to be performed every three years.

Of particular note is EPA’s renewed emphasis on the General Duty Clause of the CAA Section 112(r). This clause essentially requires regulated facilities to follow national consensus standards and standard industry practices to ensure that the consequences of a regulated chemical release are mitigated, even if such a requirement is not specifically required in the RMP regulation itself. For example, nothing in the RMP regulation requires secondary containment of a process storage vessel. However, since the use of secondary containment will likely reduce the distance to a toxic endpoint for a worst case release scenario by several orders of magnitude, it would be considered to be required under the General Duty Clause.

Another important aspect of EPA inspections is to ensure the close-out of any recommended safeguards that were made during the Hazard Review or Process Hazard Analysis (PHA). Inspectors will look to see that each recommended safeguard was implemented. If it was not, the inspector will look for documentation that justifies why a particular safeguard was not installed or implemented. Consequently, regulated facilities should review their PHAs and Hazard Reviews to ensure that all items addressed in the analysis were officially closed out and documented.

Process Weight

Reporting to EPA the correct process weight of a regulated chemical is critical. It should be noted that facilities with high process weights (e.g. in excess of 1,000,000 pounds) are considered to be “High-Risk” facilities and are targeted for more frequent inspections. Process information, including a summary of your entire RMP is required to be reported to EPA using their RMP eSubmit program on the Central Data Exchange (CDX) website.

EPA guidance dictates that the weight of a specific regulated substance in a mixture is to be used for all threshold calculation and reporting purposes, not the total weight of the mixture. However, this does not apply to aqueous ammonia. For example, 100,000 pounds of a 30 percent mixture of hydrazine (a listed RMP chemical) would equate to 30,000 pounds of hydrazine being reported as the total process weight for RMP purposes. Conversely, for aqueous ammonia solutions, the total weight of the entire mixture must be used for reporting and threshold determinations. Consequently, a facility having an aggregated total of 100,000 pounds of a 30 percent aqueous ammonia solution in its process vessels must report 100,000 pounds of aqueous ammonia in its RMP summary that is submitted to EPA.

Another common error is the omission of the total process weight of ammonia that is contained in system distribution piping. In determining this additional volume, it is important to include the amount contained in the piping system from the SCR distribution skids, strainers (if used), to the injection lances. Although this amount will likely be minimal and have little impact on the distance to a toxic endpoint in your release modeling scenarios, it nevertheless needs to be included for threshold determinations, release modeling and RMP reporting purposes.

Land Classification

Facilities have historically used the Auer Land Classification system (40 CFR Part 51 Appendix W) when performing risk assessments and air dispersion modeling scenarios for Title V air permitting calculations. It naturally follows that the Auer system should be used for determining if your facility is located in a rural or urban area for RMP purposes. While some EPA inspectors have questioned the use of the Auer system based on field observations made during the inspection process, it is the only definitive classification system for making these determinations and continues to remain the recommended method for federal air permitting purposes.

Regulated facilities should note that nine states have received delegation to implement their own RMP programs and some may have different land use classification systems. For example, South Carolina requires the use of the National Land Cover Data (NLCD) scheme for determining urban or rural land classifications for state RMP reporting purposes.

Regulated facilities located in delegated States should check with their state regulatory contact to see if they accept the Auer land use classification system.

Air Dispersion Modeling

EPA allows the use of any number of different air dispersion modeling techniques for determining the distance to a toxic endpoint. These vary in complexity from look-up tables and simplified software to more complex modeling programs such as the Aerial Locations of Hazardous Atmospheres (ALOHA), developed jointly by EPA and the National Oceanographic and Atmospheric Association (NOAA). Regulated facilities may use the modeling technique of their choice provided the rationale for its use can be justified. Results can vary widely depending on which model is used, as demonstrated below by comparing the distance to a toxic endpoint for a worst case release of 165,000 gallons of aqueous ammonia using three different modeling techniques.

As the figure illustrates, the ALOHA model provides the most conservative distance to a toxic endpoint under this scenario. The main reason for this variation is that ALOHA allows for more data inputs such as site specific meteorological data and chemical release conditions and therefore, provides a more accurate estimate of the distance to a toxic endpoint. ALOHA is also the preferred real-time modeling program used by State and Local emergency responders for use during an actual release event. Consequently, ALOHA is the preferred method to be used in RMP air dispersion modeling since it is the primary tool emergency responders will use if they have to respond to a chemical release at your facility.

When using ALOHA to model a release of aqueous ammonia, another common error is to calculate the amount of ammonia gas released from a puddle of spilled aqueous ammonia, and to then enter this release rate into ALOHA as a “direct release” of ammonia to determine how much ammonia will evaporate from the puddle. Users should be aware that ALOHA treats direct ammonia releases as anhydrous ammonia gas, which has different chemical properties than that of evaporating aqueous ammonia. Therefore, using the direct release method to model an aqueous ammonia release from a puddle spill with ALOHA will likely lead to inaccurate results.

The “puddle method” is the correct method for using ALOHA to determine the distance to a toxic endpoint for aqueous ammonia when modeling a spill puddle. This method takes into account the chemical properties and the evaporation rate of ammonia generated from a pool (diked or undiked) of spilled liquid ammonia. By using the puddle method, it is not necessary to calculate the evaporation rate of ammonia from the puddle as the program does this for you.

RMP Updating

RMPs need to be completely updated at least once every five years. Certain “trigger events” will require an update prior to five years, such as a major process change that requires an updated Offsite Consequence Analysis (OCA), an accident that resulted in or could have resulted in a catastrophic release, or even a modest change such as the name of the facility’s emergency response contact.

If you update any aspect of your RMP except for emergency contact information and accident reporting, the five-year update clock will reset to the date you updated the information on RMP eSubmit.

Regulated facilities should also be aware that a number of air dispersion modeling programs have undergone various revisions over the last few years. Even if there have been no substantive facility changes, prior modeling results calculated with earlier versions of modeling software, including ALOHA, may differ significantly from those calculated using the latest version of the software. As a result, facilities should re-evaluate their OCAs using the latest available software to ensure that the distance to a toxic endpoint has not changed significantly. If the distance has changed by a factor of two or more (doubled or halved) from your previous calculated distance, you must update the OCA information via EPA’s RMP eSubmit system within six months of this determination.

The affected receptors identified in current OCAs submitted to EPA by most facilities are based on the 2000 U.S. Census data. However, the 2010 census data is now available, documenting increasing or in some cases, decreasing populations within a given area. While regulated facilities are not required to update their RMPs just because the new census data is available, best management practices would dictate that when an update is appropriate, the newer 2010 census data be used to determine the number of receptors within the distance to a toxic endpoint.

In closing, regulated facilities should review the updated guidance documents available on EPA’s RMP website (, especially the inspection guidance document. In addition, facilities should ensure accurate accounting of all regulated process amounts in their RMP submission, re-evaluate their land use classification and review their OCAs using the latest dispersion modeling software and census data.

Expanding the Role of Natural Gas

By Chris Pais, Manager Application Engineering, FuelCell Energy, Inc

With global population now at 7 billion and climbing, demand for efficient, environmentally aware energy is going nowhere but up. According to the Energy Information Agency (EIA), the U.S. alone will see an expected 223 GW of new generating capacity demand between 2010 and 2035.

The trouble is, this increased demand can no longer be met solely by the construction of large, central generation power plants. More complex permitting for centralized plants, particularly clean air permitting, high project costs and limitations in the transmission infrastructure all point to a solution comprised of smaller power plants at locations right where the power is consumed. However, there is real opportunity for utilities, independent power producers and large industrial power users to capitalize on the natural-gas infrastructure to meet this demand by using distributed generation fuel cell power plants.

Fuel cell with heat recovery used for hot water heating at a university. Photo courtesy Fuel Cell Energy.

Natural gas is a low-carbon fuel source compared to other conventional fuels used in power generation. The average carbon dioxide emission from natural gas is approximately 119 lb/MMBtu, while that for coal ranges from 205-227 lb/MMBtu[1]. Abundant gas supplies in North America, a well-established distribution infrastructure and low natural gas costs provide an attractive long-term outlook for natural gas powered base-load distributed generation, as long as that power can be clean, quiet and efficient. To that end, fuel cells stand out as the most efficient means available to produce electricity from natural gas for their size class; low emissions and quiet operation make them particularly well-suited to distributed generation applications.

What is a fuel cell?

A fuel cell is a device that converts the energy present in a fuel to electricity using an electro-chemical process. It is similar to a battery in some ways; a battery converts stored chemical energy to electrical energy. However, a fuel cell does not store energy in the form of internally contained reactants; instead, fuel and oxidant reactions are continuously fed to a fuel cell, which converts the energy in the fuel to electricity through an electrochemical reaction between the fuel and oxygen. Like a battery, a fuel cell consists of two electrodes – an anode (which is supplied with fuel) and a cathode (which is supplied with oxygen, typically ambient air). An electrolyte separates the electrodes and conducts ions between the anode and the cathode, which then drives an external electrical load, as shown in the figure.

Fuel cells require hydrogen as the fuel source. However, due to the absence of a well-established hydrogen infrastructure, commercial fuel cell manufacturers make provisions in their equipment for producing hydrogen from a readily available fuel source such as natural gas, or renewable biogas. In low temperature fuel cells, the hydrogen conversion system is external to the fuel cell. In high temperature fuel cells, the hydrogen conversion system is integral to the fuel cell, resulting in higher efficiencies.

Types of Stationary Fuel Cells

There are four types of commercially available stationary fuel cells, and they are named after the type of material used in the electrolyte. The figure shows the different type of fuel cells and their associated chemical reactions.

Table 1 summarizes the attributes of commercially available fuel cells. Fuel cells commercially deployed in large stationary power applications typically include phosphoric acid, molten carbonate and solid oxide types, and range in power output from 100 kW to an 11.2 MW fuel cell park, the world’s largest as of the date of this article.

Fuel Cells are Clean and Efficient

Unlike conventional power plants, fuel cells extract electricity from the fuel without going through a combustion process. This results in a clean emission signature, without any of the pollutants associated with combustion such as nitrogen oxides (NOx) or particulate matter. Table 2 compares the typical pollutants generated by fuel cells with those generated by conventional methods, illustrating how fuel cells virtually eliminate pollutants from the power generation process and significantly reduce greenhouse gas emissions versus the average U.S. fossil fuel power plant.

Due to the direct conversion to electricity, fuel cells are also more electrically efficient than conventional combustion, especially in the distributed generation size range. In fact, fuel cells yield more power per a given unit of fuel than virtually all other methods of power production. They typically range from 40 to 60 percent electrical efficiency, and can achieve up to 90 percent total efficiency in Combined Heat and Power (CHP) applications.

Combined Heat and Power from Fuel Cells

One of the more advantageous byproducts of the fuel cell reaction is heat, which can then be used in a CHP setting to increase the total efficiency of the system. High temperature fuel cells, such as molten carbonate (MCFCs) lend themselves particularly well for heat recovery due to the high temperature (>700 F) of the exhaust gas. Users can put this high temperature exhaust to use for a variety of value-add applications such as hot water heating, steam generation or absorption cooling.


Stationary fuel cell power plants fueled by natural gas are easy to site in urban locations due to their favorable emission profile, relatively modest space requirements, and quiet operation. Costs continue to decline as manufacturing volumes increase, although comparing current direct cost per kWh without accounting for attributes such as the virtual lack of pollutants does puts fuel cells at a pricepoint that is more than combustion-based generation equipment at present. However, in certain high cost regions such as California and Connecticut, carbonate fuel cell plants are close to being competitive with the grid, even before incentives. Fuel cell manufacturers have a clear path for achieving pricing below the grid as higher manufacturing volumes will further drive costs down.

A 11.2 MW fuel cell installation in South Korea with thermal energy sold to a neighboring wastewater treatment facility

Recent advances in fuel cell technology, increases in production volume and competition among manufacturers of stationary fuel cell systems continue to lower the gross system costs. Various states in the U.S. have attractive programs for the deployment of fuel cells. For example, the State of California renewed the Self Generation Incentive Program (SGIP) in November 2011, which offers a capital cost incentive of up to $2,250 per kW for natural gas and $4,250 per kW for biogas fueled fuel cells. California also recently enacted a feed-in tariff for CHP applications. Connecticut has a similar capital cost incentive program for fuel cells, as well as an attractive feed-in-tariff scheme. In addition to several state incentives, the Federal Investment Tax Credit (ITC) provides a tax credit of 30 percent of the project costs, up to $3,000/kW for fuel cell plants, and the Energy Policy Act of 2005 still provides for accelerated depreciation of fuel cell assets for tax purposes.

In the international arena, South Korea includes fuel cells operating on either natural gas or renewable biogas in its Renewable Portfolio Standards (RPS) and has an attractive feed-in-tariff program for fuel cells. The high efficiency of fuel cells is especially attractive to South Korea, as it imports about 95 percent of the natural gas used for power production. European countries also provide an array of capital cost and operating incentives for combined heat and power applications, with favorable treatment for fuel cell generation in some countries.


The long-term abundance and low cost of natural gas provides real, long-term opportunity for power producers and large-scale consumers, especially as fuel cell technology continues to come into its own. Public opinion and policymakers are embracing efficiency and environmentally friendly power production while moving away from combustion-based processes and nuclear power. Today, there are several manufacturers of fuel cells for industrial-grade, stationary power applications, ranging in size from 100 kW to several megawatts. Fuel cells are characterized by high efficiency, low emissions, low-noise and can be easily sited in urban areas due to these attributes. Some even provide the advantage of heat recovery, further enhancing the technology’s economic and environmental value for distributed, stationary base-load generation.


1. Table 1, Voluntary Reporting of Greenhouse Gases Program Fuel Emissions Coefficients, EIA,

2. Pais C and Leo A, Combined Heat and Power Applications of High Temperature Fuel Cells, Cogeneration and Distributed Generation Journal, Volume 25, Number 3, pp 7-25, Summer 2010.

3. Values for US Grid and non fuel cell generators from “Model Regulations for the Output of Specified Air Emissions from Smaller-Scale Electric Generation Resources”, The Regulatory Assistance Project, for the National Renewable Energy Laboratory (NREL); October 15, 2002. Values for NOx, SOx and CO2 from US EPA eGRID 2007, version 1, Year 2005 Summary Tables.

More Power Engineering Issue Articles
Power Engineerng Issue Archives
View Power Generation Articles on