By Robert L. Branning and Michael J. Pelzer
From 1997 through 2004, SJRPP’s fuel supply consisted of 80% coal and 20% petroleum coke. The coal is supplied under two separate long-term contracts and the petroleum coke is purchased on the spot market. Due to the fixed price of the contract coal, SJRPP’s aggregate fuel cost remained largely unaffected by rising fuel prices in 2003 and 2004. Only the 20% component representing petroleum coke was affected by market forces. Still, there remained a distinct advantage in burning petroleum coke as the differential price between petroleum coke and coal remained substantial. Due to the combustion improvements on Units 1 and 2 in 2003 and 2004 respectively, both units began burning 30% petroleum coke in April, 2005.
Looking forward from 2004, the first of the two long-term coal contracts was set for a market re-opener in December, 2005. Because the low heating value of this coal challenged the fuel delivery system, a substantial increase in fuel cost was anticipated beyond 2005. To avoid or mitigate this expected fuel cost increase would require SJRPP to burn a lower-grade coal than the coal which had previously defined their lower limit of coal quality (11,800 BTU/#).
This coal supplier suggested a lower-grade coal averaging 11,300 BTU/# at a substantial discount. Evaluating the statistical variance in the coal’s heat content, a 7 percent increase in pulverizer coal throughput would be required to make this substitute coal viable.
The OEM mills are vertical spindle pulverizers with three static rotating grinding rollers and a rotary grinding zone. They were designed to be a low-speed, high-volume mill with design capacity of 103.8 K#/hr at 70% fineness passing 200 mesh while grinding a 48 HGI (Hardgrove Grindibility Index) coal. However, the mills had never performed at this level and the base capacity of 89.5 K#/hr (70% through 200 mesh, 48 HGI) for these mills was established by testing. This base capacity and standard correction curves have historically been used for screening candidate fuels.
The OEM mills had originally been furnished with stationary air ports. To reduce maintenance costs, these were later replaced by aftermarket rotating air ports which resulted in no effect on mill capacity. The OEM classifiers were designed to be externally adjustable, but the mechanism proved inoperable since start-up and required a mill shutdown to adjust. The OEM discharge rings and classifier reject chutes were original. The grinding ring segments were OEM castings and the grinding rollers were aftermarket welded overlay material and included a spindle design modification. While the aftermarket modifications were successful in reducing maintenance costs, they had no effect on mill capacity. Pulverizer capacity remained the obstacle to burning lower rank, less expensive fuels.
Figure 1 - SJRPP Pulverizer Prior to SAS Modifications
PULVERIZER PERFORMANCE EVALUATION
The problems revealed during the SAS Global Power evaluation were relatively common: restricted capacity due to air port and classifier limitations. The main constraint to reducing fuel costs was mill capacity. Above its maximum capacity, the mill differential and motor amps would gradually increase as the feed rate exceeded the mill’s capacity to grind and fuel inventory within the mill increased. As mill table differential pressure increased, it would fluctuate as the mass of fuel exiting the pulverizer cycled, creating thermal upsets boiler control. This would often require the operators to keep the pulverizer’s feeder and air control in manual mode.
Another issue involved the classifier reject discharge chutes. The weighted trap doors would tend to bind or lock in a throttled position reducing the return of oversized, rejected particles to the grinding zone. As additional chutes became bound, the classifier capacity would diminish as the path for oversized fuel returning to the grinding zone became restricted. Ultimately, the classifier cone would fill with fuel, forcing oversized fuel particles to exit the mill prematurely. This resulted in reduced fineness, poor flame quality, high carbon monoxide production and high LOI (Loss On Ignition, or unburned carbon in ash). Correcting the problem required personnel entry which rendered the mill unavailable for a day. This problem was exacerbated by wet fuel conditions as the elevated primary air temperature would cake or char the contents of the classifier into a solid mass.
In Spring of 2004, SAS built and SJRPP installed a prototypical Rotating Air Port in the 1B Mill. This mill was selected as the prototype since it had the history of being the worst of the fourteen mills with regard to mill bowl differential pressure and the aforementioned “surging.” Consequently, this mill historically had the lowest fuel loading and required the remaining mills to make up for its deficiency.
Upon completion of the air port installation, results were immediately observed as the operators increased the participation of the 1B Mill. The effectiveness of the SAS Rotating Air Port to reduce classifier duty through primary classification resulted in additional capacity from the mill. Based on this success, it was determined to prototype an additional mill with all the features SAS recommended to improve capacity and performance.
ADJUSTABLE SAS ROTATING AIR PORT
The Adjustable SAS Global Patented Rotating Air Port design, Figure 2, is the result of many years of R & D, CFD modeling and extensive experience. The air directional vanes of the rotating air port face the direction of grinding table rotation at a precise angle. Combined with the throat annulus area, these vector the air port velocity at approximately 10,200 fpm. The vanes are also adjustable to accommodate fuel and/or operating changes. Through the rotation of the air port with the grinding table, the primary air circulates circumferentially -- an efficient fuel drying method. The cantilever design of the vanes allows the radial seal to utilize the total air port area. The deflectors, immediately above the vanes, move the coal/air mixture toward the mill’s center providing primary classification of the coal and depositing larger particle sizes back onto the grinding segments. The 50 mesh fineness levels, which are largely responsible for LOI, are significantly improved at this mill level and the reduction in classifier turndown contributes to the grinding capacity increase.
Figure 2 - Adjustable SAS Rotating Air Port
HIGH SPIN STATIC CLASSIFIER
A SAS Global Patented High Spin Static Classifier Blade, Cage and Adjustable Cycling Mechanism, Figure 3, replaced the OEM Cage Assembly. The blades are 180 degrees from the OEM arrangement and cycle around a pivot point for blade adjustment to the discharge ring tangent. There is a downward deflection in the blade form and the blades extend downward below the classifier cone top. This design takes advantage of the rotational forces imparted on the dense phase flow within the pulverizer at the primary air porting elevation. The conical form, depth and direction of the blade, accelerate the circumferential movement of the dense phase flow around the classifier cone which acts as a cyclone separator. Centrifugal force controls the fineness level and the blades can be easily rotated from the mill’s exterior. The original adjustment mechanism was completely redesigned utilizing stainless steel gimball bushings, turnbuckles and pivot arms. The original actuation gearbox and connecting arm were re-used. This device provides control of the 200 mesh fineness as well as classifier differential and aids in lowering system resistance through the elimination of restriction. The classifier cone was replaced and built in three segments to facilitate any future maintenance requirements.
Figure 3 - High Spin Static Classifier
SAS CONE EXTENSION
The SAS Global Patented Cone Extension Kit, Figure 4, eliminates the OEM classifier rejects discharge chutes, thus eliminating the mechanical trap doors and their potential to bind. The feed pipe is extended lower in the mill and a cone extension is bolted to the bottom of the classifier cone, extending it approximately 48 inches. A series of pipe spools are then bolted to the bottom of the classifier cone extension. This classifier component returns the oversized, rejected particles back into the grinding zone, approximately 32 inches above the table. By eliminating the discharge doors and returning the fuel directly into the grinding zone, mill differential can be reduced, fineness improved and plugging or “surging” is eliminated. In addition, service life of the classifier components is extended.
Figure 4 - SAS Cone Extension
SAS MULTI-OUTLET DIFFUSER
The SAS Patented Multi-Outlet Diffuser (MOD) creates a homogenous mixture of fuel and air in the turret section of the pulverizer. This allows a balanced fuel flow thru each of the pipes exiting the mill. The stimulation of flow characteristics enhances particle size distribution, dense phase flow mixing and coal pipe fuel distribution. Fuel distribution is critical to the effective reduction of NOx and LOI concurrently.
Figure 5 - SAS Multi-Outlet Diffuser
EVALUATION OF COMPLETE SAS UPGRADE
The 2G Mill was selected to receive the complete SAS upgrade due to its grinding surfaces having been recently restored in August 2004. Baseline testing was performed in September, 2004. After a brief delay due to the active hurricane season, the upgrade was performed in October 2004 and the mill re-tested in November, 2004.
For these tests, the mill output was slowly raised until its maximum capacity was established. Additional fuel input beyond this point would prohibit steady state operation as observed from an unending increasing mill bowl differential pressure. This would ultimately lead to the surging effects previously mentioned. All tests were performed with coal from the lower grade supplier.
The baseline testing in September 2004 with restored grinding surfaces and prior to the SAS upgrade, established the mill’s baseline capacity of 90.2 K#/hr when corrected to standard fineness of 70% through 200 mesh. These test conditions are presented in Table 1 and mill capacity corrected to standard fineness is shown in Figure 6.
This result fared favorably with the previously established base capacity of 89.5 K#/hr that had been the historic basis for fuel procurement.
Table 1 - 2G Mill Capacity Testing Before and After SAS Upgrade
Upon completing the upgrade in October 2004, the mill was returned to service and re-tested in November 2004. At the onset of the re-testing, it became immediately apparent that the mill’s new capacity exceeded the 100 k#/hr capacity of the coal feeder. The feeders were re-configured and testing resumed.
Since this was the first test performed since the upgrade, the mill’s fineness was unknown at the time of the test. The classifier was in a “neutral” position based on SAS’s experience and no problems were noted with flame stability, CO or LOI. Test conditions from the first test following the complete SAS upgrade are provided in Table 1 and mill capacity corrected to standard fineness is shown in Figure 6.
Following the first test, the classifier was adjusted to produce a finer product. This adjustment was made with the mill in service while monitoring mill motor amps, feeder speed and differential pressure to ensure the desired classifier position. Test conditions from this test are provided in Table 1 and mill capacity corrected to standard fineness is shown in Figure 6.
As shown in Table 1, fineness from the second test is greatly improved and the corrected capacity confirms the capacity increase observed during the first test. The two tests demonstrated an average corrected capacity of 103.0 K#/hr after the complete SAS upgrade versus 90.5 K#/hr prior to the upgrade. The 14% capacity increase was double SJRPP’s requirement of a 7% increase for consideration of using the lower quality coal.
Figure 6 - Comparison of Corrected Mill Capacity
Before and After SAS Upgrade
Based on these results and favorable operating experience from October, 2004 to April, 2005, SJRPP accepted the 11,300 BTU/# offer in lieu of the December, 2005 market re-opener. This mitigated the cost increase expected from the re-opener by $6.3 million annually beginning in 2006.
Additionally, in April, 2005, SJRPP placed an order with SAS to upgrade the remaining fleet of 13 mills and an aggressive schedule was developed for implementation. This allowed SJRPP to negotiate substituting the lower grade coal for the remaining 2005 contractual quantity of higher grade coal for a fuel savings of over $2.5 million in 2005. In January of 2005, SJRPP received their last shipment of the higher grade 11,800 BTU/# from this supplier.
Pulverizer capacity at SJRPP has historically limited fuel supply to coals at or above 11,800 BTU/#. The pending expiration of long-term coal contracts created the potential for substantial fuel cost increases.
SAS’s Comprehensive Systemic Evaluation of the mills at SJRPP correctly identified the pulverizer classifier as the mill’s limitation. Through an improved rotating air port design, classifier duty was reduced utilizing primary gravitational classification. Additionally, an improved classifier reject chutes design eliminated restrictions created by the original trap door design. This change, along with a less restrictive classifier cage, and more aerodynamic classifier vanes have resulted in a 14% increase in mill capacity.
As a result, SJRPP has expanded its coal market, successfully renegotiated one of its two long-term coal supply contracts, and positioned itself more favorably for future fuel procurement.
SJRPP has just completed a 3 month long trial burn using an 11,000 Btu Coal. The Unit was able to reach full load very easily without having to max out any of the 6 mills. In fact, the test burn went so well, SJRPP has decided to start using a 10,500 Btu fuel. SJRPP has also stopped burning Petcoke, due to the fact the low Btu Coal they are currently using is much more cost effective.
1. Branning, R., Conn, R. and Vatsky, J.; “Reducing NOx and LOI at the St. Johns River Power Park”; Power-Gen International 2003; December 9-11, 2003; Las Vegas, NV
2. Branning, R., Conn, R., Schiazza, G. and Vatsky, J.; “Firing 100% Petroleum Coke in Burners at the St. Johns River Power Park”; Clearwater Coal Conference; April, 2005; Clearwater, FL.