Reaching emissions and performance goals often requires that we think creatively and be open to new approaches and technologies. That’s been our philosophy at CPS Energy. We are the nation’s first power generator to integrate combustion optimization software with boiler optimization spectroscopy sensors. The combination of technologies is new and we’re still in the project’s early stages, but the results we’re seeing so far are encouraging.

Regulation trends and the inevitability of SCR systems
In 1999, CPS Energy management could see the writing on the wall and believed that eventually, all coal-fired units would have to install SCRs or SNCRs to meet NOx emissions requirements. So we prioritized getting NOx as low as we possibly could. If we could get this emission low enough in the furnace, it would mean smaller SCRs, and lower associated capital and O&M costs.

We had great success with in-furnace NOx reduction by ensuring controls were properly tuned, staging of secondary air in the main windboxes and installation of low NOx nozzle tips. We also tried balancing air flow in the coal pipes with practically no success, so we gave up on that.

As part of our commitment to maximize in-furnace reduction of NOX emissions, we also installed CombustionOpt® on our three coal fired units: Spruce 1, and Deely 1 and 2. The software lowered NOx further.

How low can we go?
Although we had made progress lowering NOx, we believed that we could do even better. We were realizing between 0.12 and 0.13 lbs/mmbtu but had set a goal of 0.08 lb/mmbtu. We were reading CO, O2 and temperature metrics from the normal location – the economizer outlet – and feeding that data to the combustion optimizers. The problem is that this is not a real-time reading and, by the time flue gas reaches the economizer it’s already been corrupted by tramp air. I knew that if we could get accurate readings just above the fireball, then we could feed real-time and more precise data to CombustionOpt and optimize even more aggressively.

When I heard about the ZoloBOSS™ system it seemed like what I had been looking for: a grid of lasers that provide clean, real-time readings of boiler conditions, including CO, O2, CO2, temperature and H2O. I’ve always felt that if these constituents could be balanced across the furnace, NOx emissions and heat rate could be greatly improved. Zolo’s proposition was highly scientific, but my strongest area of knowledge is boilers. I was interested to hear what my more techie colleagues at NeuCo thought about the ZoloBOSS system. So after the 2002 Coal-Gen Conference, I facilitated a meeting in San Antonio, TX between my colleagues at NeuCo and my new contacts at Zolo to talk shop and see if combining these two technologies could strengthen the performance of both. After much discussion, I left the meeting excited about the possibilities of combining these solutions to lower NOx even further at JT Deely. 

Integration and going closed-loop
We presented the concept to CPSE management. They were very receptive and supportive, and gave us the go-ahead to begin installation and integration on JT Deely 1. On July 7, 2008, we went into closed-loop operation, with the ZoloBOSS system feeding its highly accurate data to CombustionOpt, which in turn is able to optimize more aggressively for NOx and heat rate reduction. So far, we’ve seen improvements on NOx and heat rate, but we’d like to see those improvements on a more constant basis. Presently, we’re concentrating on balancing O2 across the furnace. As the O2 is balanced, we’ve been able to reduce excess air levels to reduce NOx emissions with no evidence of an increase in reducing atmosphere regions (slagging) or any other ill effects to the boiler. At the same time we’re working to allow all windbox dampers to work independently of one another. Independent operation of the dampers will open an entirely new and wider field for CombustionOpt. I’m confident that we can reach a constant 0.09 lb/mmbtu in ‘09, and then I think we can set our sights on lowering NOx even further. Greedy? Perhaps. But why not?