The combination of a reasonably large ‘spark spread’ and relatively low natural gas prices for now and the future helps to make the business case for cogeneration schemes at industrial and commercial sites in the US, according to analysis by David C Oehl at Maven Power.
With issues related to sustainable development pushing actions towards improving electrical generation efficiency, the case for industrial-scale cogeneration has become more compelling. Cogeneration, the simultan-eous generation of electrical power and heat, usually in the form of steam or hot water, is increasingly becoming a viable option for both domestic and international industrial plants. Opportune industries include pulp and paper; breweries; bottling and canneries; manufacturing; agricultural mills; steel; chemical; cement; and building sectors: hospitals and university campuses.
The increased viability is due to consistently low natural gas costs and electricity prices resistant to falling in step with generation fuel prices. Moreover, gas prices are expected to remain at historic lows for some time to come in the US as the country currently sits on ample reserves for the next 120 years and as a result of a growing aversion to imported foreign energy sources. In addition, interest rates and companies’ cost of capital remain extremely low, further pushing the cost of new projects or repowering efforts lower.
|Figure 1. Project payback versus electricity price|
With traditional renewable energy technologies, such as solar photovoltaic and wind energy, consistently unable to prove financially or physically accessible to large populations of the country at any reasonable scale, natural gas, as the cleanest of all fossil fuels and more than twice as clean as coal, will continue to be the obvious choice for industrial on-site generation – with industrial cogeneration as an attractive long term option.
Typical industrial or large commercial loads fitting the current target profile include plants or facilities having electrical loads in the 5–20 MW range and steam loads of 6800–3600 kg/hr.
INDUSTRIAL FEASIBILITY STUDY
Consider a pre-engineering and modelling feasibility study for an industrial cogeneration power plant performed by Maven Power of Houston, Texas. To illustrate the most challenging economic case, an application at the lower end of the power and steam load range was analyzed. The study was based on an industrial plant requiring 5.3 MW of electrical power and two steam conditions for the plant processes.
The objective of the study was to determine the techno-economic feasibility of on-site self-generation of power and steam using a turbine-based cogeneration plant versus purchasing utility electrical power and steam generation using traditional on-site boilers.
The cogeneration power plant was based on a single Solar Taurus 60 gas turbine generator and accompanying heat recovery steam generator (HRSG) located in the United States. The gas turbine was modelled using the manufacturer’s SoLoNOx dry low emissions (DLE) technology, but selective catalytic reduction (SCR, NOx reduction only, no CO catalyst reduction included) equipment was included in the modelling to ensure the plant would qualify as a minor source of emissions as defined by some regulating authorities.
The study yielded the following performance results (performance based on continuous power output at 92.5% capacity factor and 8100 hours per year):
• power output: 5306 kW
• steam output: 11,000 kg/hr
• CHP efficiency: 81.9%
Maven Power’s modelling yielded the following results as related to the base line green-field site considerations:
• expected water usage (assuming all process steam consumed by customer’s process, with none returning as condensate): 12,100 litres/hr at 24ºC;
• fuel consumption: 1350 kg/hr natural gas (17.3 MW/hr)
• required site area: 31 x 34 metres;
• emissions: NOx, 4.40 tonnes/yr (as NO2); CO, 26.8 tonnes/yr; CO2, 28,400 tonnes/yr;
• ammonia consumption (SCR): pure (NH3), 6.5 tonnes/yr; aqueous, 22.4 tonnes/yr.
The economic feasibility of this 5.3 MW application was modelled and Figure 1 shows a representative case for time to payback versus electricity price for a range of natural gas prices ($4–$14).
As an example, if an industrial user’s average power price is $0.10/kWh and the fuel price is $6/MMBtu ($20.5/MWh), time to payback is 3.4 years. As can be seen, payback times are highly attractive in today’s market for fuel prices lower than $8 and electricity prices above $0.09/kWh.
A new figure can be generated for each application’s plant size in which the payback time gets shorter with increasing plant size.
|Rice University’s Centaur 50 with heat recovery|
UNIVERSITY CASE STUDY
Consider Rice University’s campus of nearly 6000 students located in Houston, Texas. Rice’s central energy plant is located in the centre of campus and serves to provide the university’s electric power and district cooling and heating needs. The plant currently employs two gas turbines: a primary unit, which is a Solar Centaur 50, and a standby unit, a Ruston TB5000, both of which are equipped with heat recovery steam generation units to meet their electrical and steam demands.
Power demand require-ments for the campus range from 10 to 17 MW, with up to 4 MW normally being supplied by the Centaur 50 gas turbine and the balance being made up by the local utility. Steam demand is typically flat during the course of the year and in the range of 7260–9980 kg/hr. It is normally met by the Centaur 50’s waste heat recovery unit. During steam load peaks, which can reach up to 13,610 kg/hr usually needed for district heating during cold weather, the additional steam load is met by the central plant’s on-site gas-fired boilers.
The boilers are not normally used for operation unless the demand requires it, since the steam generation from the turbine waste heat is far more economical. Both electrical power and steam generation are provided to the campus 24 hours a day, seven days a week.
These operating conditions place Rice at the lower end of the target range for what the Maven Power study identified as attractive – small-scale cogeneration plants. According to Doug Wells, Rice’s Central Plant Director, Rice has considered a repowering effort in recent years, based on expected student growth and ageing existing assets.
One option considered was the installation of a new Solar Mercury 50 with heat recovery, which would add approximately 4.0 MW of electrical generation capacity and 5760 kg/hr of steam generation. Based on Rice’s current utility costs of $0.10–0.11/kWh for electricity and $0.020–0.027/kWh for natural gas, referring to Figure 1, it can be seen that a payback could be realized in 3.5 to 4.5 years. Accounting for the fact that the plant size is slightly lower at 4.0 MW, the payment is extended slightly, yielding a 4.5–5.5 year payback period.
The Mercury option was initially attractive due to its excellent simple cycle efficiencies (which would be in the 37% range) and the fact that it wouldn’t require removal of existing equipment. Based on Rice’s load profile and its currently-installed power and steam generation equipment, it is reasonable to give priority to a lower cost electricity generation solution which is also able to meet the entire steam load.
|An on-site gas-fired boiler|
However, this option is problematic for two reasons. First, the Mercury causes the overall CHP efficiency to fall dramatically due to its lower level of available waste heat and, secondly, the amount of steam generated (3490 kg/hr) is well short of the university’s demand requirement. The object, given the situation at Rice, is to find the highest efficiency turbine unit which can simultaneously generate waste heat to closely match the average steam load.
Given an impending repowering situation in which Rice’s assets would be replaced, it is worth considering the cogeneration model as given by the Maven Power study. The 5.3 MW Taurus 60 unit, which could generate 9070–11,340 kg/hr of steam, would be a close match for the university’s current steam demand and allow for reasonable steam load expansion. Moreover, this option dramatically improves the CHP efficiency over the Mercury option and is superior to the current Centaur50 due to its higher simple cycle efficiency and greater steam generation capacity. The higher power output of the Taurus unit may provide an opportunity to lower utility demand or time of use charges as well. Again, referring to Figure 1, payback based on the university’s utility costs could occur in five years or less.
In the current market, given the reasonably large ‘spark spread’ between electricity and fuel costs, and the expectation for natural gas prices to remain suppressed for the foreseeable future, small-scale cogeneration in industrial and other applications is increasingly attractive.
Moreover, even with longer-term fuel price volatility an uncertainty, with short break-even payback periods as demonstrated in the Maven Power study and the Rice University case analysis, risk is significantly reduced to the owner or end user. Further arguing the case, the study presented focuses on a near worst case scenario in terms of scaling – a single turbine/HRSG configuration generating relatively small amounts of power and steam. The economics and overall risk are significantly improved for users with greater electric power and steam loads.
David C Oehl is President, Maven Power, Houston, Texas, US. Email: firstname.lastname@example.org
This article is available on-line. Please visit www.cospp.com