YARDLEY FLOODING (Q&A)
Brock Creek below the Canal Bridge Aqueduct
By Larry Hale - March 10, 2005
A homeowner's question mailed to me on February 22, 2005 relates to flooding below the Canal Aqueduct and whether or not it would be possible to protect the houses on East Afton by constructing a wall along Brock Creek. Homeowners observe that floodwaters always arrive from the direction of Brock Creek, either directly from creek flooding, or indirectly from the river by backing up under the bridge. Would it be possible to control this by constructing a wall along Brock Creek? The short answer is...
the extent to which a wall along Brock Creek would help depends on its height but it would only protect up to the level at which waters started arriving from a different direction and would not protect against a 100-yr flood at which level the entire area is covered and the water is coming from all directions
Your question concerns both river and creek flooding in a specific area along Brock Creek between the canal and the river. The first part of my analysis focuses on the river, the second part on the canal. Many of my comments may be applicable to the overall Yardley area between the canal and the river.
Reference:
"Pennsylvania's Delaware Division Canal: Sixty Miles of Euphoria and Frustration" by Albright G. Zimmerman, Published by: Canal History and Technology Press, 2002 [Hugh Moore Historical Park and Museum, Inc., 30 Centre Square, Easton, PA 18042-7743]
The above publication is of particular value because it provides extensive detailed information about the operation and maintenance of the canal and problems encountered during the 100-years the canal was fully operational. Most of these problems related to water levels and flooding, and these are exactly the same problems which remain with us today!
RIVER FLOODING
Looking at a map you can see that the Delaware River is aimed straight at Yardley. If you draw a straight line down the center of the river it intersects the edge of the river at the Borough Line, crosses Morgan Avenue, and passes through the canal aqueduct ---whereas the actual river channel bends 40 degrees to the left. Long ago when the glacier was melting the river was much higher and wider. As the river subsided it left a broad sand-and-gravel bar forming an island starting at the Yardley Inn with a high ridgeline on North Bell Ave. extending all the way down through the Macclesfield subdivision. The river split into two channels, and during floods it tries to do the same thing again.
How high any flood rises depends many factors: 1) Amount and intensity of rainfall throughout the watershed. 2) Geometry of the watershed, determines how much of the runoff arrives at the same time. 3) Ground absorption. 4) Surface water retention by natural or man-made structures. 5) Downstream backups, blockages caused by ice or debris. 6) Upstream releases, controlled or otherwise. 7) Melting snow. 8) Preexisting or developing high and low pressure weather fronts.
Flood levels are described statistically in terms of frequency of occurrence, and in these terms almost any flood level is possible. Let me explain this in more detail as follows. Although the arithmetic is very precise and data collection has greatly improved, trying to describe river flooding this way is limited because it's based on what has happened in the past over a limited period for which accurate records are available. There's always a chance that some unpredictable combination of factors might occur simultaneously in the watershed. A rare combination of conditions occurred in 1955 when there were two hurricanes one closely following the other on the same overland path ---plus, a high pressure front (moving down from Canada) stalled or slowed down the sideways movement of the second storm causing it to dump 8-10-12 inches of rain on the already saturated ground in the upper watershed. It's not a question of global warming as you guessed, but rather that combinations never seen before might occur at almost any time. Furthermore, no one in the valley should assume they are immune to flooding just because they are above the 100-yr flood zone.
As a practical matter, a 100-yr flood line has somehow been drawn on a map. And today, insurance companies, federal agencies, and state agencies have mutually (all across the United States) adopted the 100-year flood as a design standard. And once they set the 100-yr target, i.e. a flood that occurs on average once every 100 years ---once that flood level was established, they will not normally fund a project to achieve protection for anything less severe. For this reason it is unlikely that a wall along Brock Creek would be funded.
Even though new construction has to be set above the 100-yr flood elevation, owners of existing homes are of course allowed to take lesser measures to protect their homes. Bear in mind that ground elevations in Yardley's 100-yr flood zone vary widely. If you are relatively high along the ridgeline of what use to be the island, many homeowners probably don't do anything. But if your are only a short distance above normal high river level, your options may be limited to either raising the house or constructing an extension from an upper level, and/or of course obtaining flood insurance. The ground surface of structures located at various intermediate elevations between the canal and the river varies roughly from more than 2-ft height above the canal towpath, to as much as 8-ft under the towpath. Differences up to 10-ft are very significant in terms of flooding, and options such as flood proofing will depend on the particular situation.
CANAL FLOODING
Creeks Flowing into Canal:C.P. Yoder in his book on the Delaware Canal said the following…
"In addition to the streams that flow under the canal by means of culverts and aqueducts, about twenty creeks flow directly into the canal channel. These creeks, however, contribute little to the water supply because in summer, when water is most needed, they are practically dry. Their main contribution, which added to the problems of the maintenance men, was to deposit large quantities of silt into the channel during the spring runoff or when heavy storms caused freshets."
[Ref: "Delaware Canal Journal" by C.P. Yoder, Canal Press, 1972]
The questionable value of these creeks during the dry months isn't our concern here. We're only focused on the fact that the canal system was intended to handle the spring runoff and heavy storm flow.
Before the canal was built, the proposed site of the aqueduct in Yardley and the design concept between Lock 6 and Lock 5 must have looked very good indeed. It provided a classic two-creek solution. Brock Creek was centrally located and had a large cross section on the upstream side and a wide floodway starting 200-ft on the side toward the river. Silver Creek is located nearby at the same elevation as the canal so that it can serve as a natural feeder ---it's watershed is less than one-third the size of Brock Creek. Most importantly, Brock Creek is so deep under the canal that the bridge aqueduct should easily handle and discharge to Brock Creek, almost any excess flow that is delivered to it by flooding from Silver Creek. The concept is still standard today... one creek supplies make-up water; the other creek removes excess water.
Waste Weir:The following was written by a canal expert in 1911...
"An essential adjunct to a canal is a sufficient number of waste-weirs to discharge surplus water accumulating during floods, which, if not provided with an exit, may overflow the tow-path, and cause a breach in the banks, stoppage of the traffic, and damage to adjoining lands. The number and positions of these waste-weirs must depend on the nature of the country through which the canal passes. Wherever the canal crosses a stream a waste-weir should be formed in the aqueduct; but independently of this the engineer must consider at what points large influxes of water may be apprehended, and must at such places form not only waste-weirs of sufficient size to carry off the surplus, but also artificial courses for its discharge into the nearest streams. These waste-weirs are placed at the top water-level of the canal, so that when a flood occurs the water flows over them and thus relieves the banks."
Ref: "The Encyclopedia Britannica" 1911 [Miniaturized Edition - Author's Library].
Specifically, the waste weir was expected to discharge excess water from flooding in the watershed serving the feeder. The design included all the necessary components: 1) The natural feeder augments waters arriving from the upper canal; 2) The aqueduct bridge prevents flood waters backing up Brock Creek; 3) The waste weir maintains the level and handles flooding from the feeder; 4) The emergency sluice gate is provided at Brock Creek. The Canal Bridge Aqueduct location represented an easy design because there is plenty of elevation head over Brock Creek. The usual bypass was provided at Lock 5.
Height of Towpath:In 1857 Edward F. Gay, newly appointed state engineer, made a detailed survey of the conditions along the canal and included the following statement in his report...
"The surveys made during the past season, for the extension of the Improvements from Frenchtown to New Hope, and thence to Bristol proves the Towing path to be lower, and the depth of water in the Canal less than had been expected. The former averages but a little over one foot over the water surface, while the latter does not exceed four feet eight inches in depth on the Lock mitre sills. I have no doubt that all Calculations heretofore made, in reference to this improvement, were based on the supposition that the depth of water was full five feet, and the bank on an average of at least 18 inches above the water surface as on all new Canals, a much larger amount of work will be necessary."
[Ref: "Pennsylvania's Delaware Division Canal" by Albright G. Zimmerman]
The crest of the weir on the two sides of the aqueduct at Brock Creek is presumably set 5-ft+ above the bottom of the canal. The towpath at the aqueduct bridge measures about 20-inches above the aqueduct weir. My interpretation of Mr. Gay's phrase "average of at least 18 inches" is that it refers to locations where the original ground surface was required to be filled under the contract and that significant stretches of the original land which was higher than the required 18-inches are not included in the calculation, and therefore, that the top of the towpath in Yardley was intended to be 18-inches above the weir, within some specified tolerance.
Depth of Water Above Weir vs. Height of Towpath:During the 1850s the Canal Company lamented…
"In regard to this portion of the State improvements, we are obliged to reiterate the oft repeated statement of insufficient depth of water - of contracted and one self constructed aqueduct and locks - of banks and towing paths inadequately supplied with overfalls, imperfectly and partially protected by slope walls and by paving and so little raised above the surface of the water in the levels, as to be liable to be overflowed and breached by every passing summer shower;"
[Ref: "Pennsylvania's Delaware Division Canal" by Albright G. Zimmerman]
At the start of any storm, there should be available at the aqueduct weir 1-ft or more (plus freeboard) of elevation head ---the height of the water above the weir is what determines the amount of discharge over the weir to Brock Creek. A chain is only as strong as its weakest link. So if at any location along the towpath the elevation is less than a foot or more above the weir, then the capacity of the weir at the aqueduct will be correspondingly reduced. There are 3-5 locations between Locks 6 and 5 where the canal sometimes overflows and this may be the reason why the aqueduct overflow weirs alone don't automatically prevent canal overflow from even relatively minor creek flooding, without the use of the sluce gate.
Trouble at the Canal Aqueduct Bridge:Problems with the canal were highlighted (in Yardleyville) in 1868...
"During the year 1868, a number of interruptions to navigation which were characteristic of others to follow, occurred on the Delaware Canal... In June of that year, a freshet in Buck Creek [Brock Creek] Swept away the aqueduct in Yardley."
[Ref: "Delaware Canal Journal" by C.P. Yoder, Canal Press, 1972]
On June 12, 1996 a rare storm occurred over local watersheds and produced a torrential downpour causing the canal to overflow along much of its length between Lock 6 and Lock 5. The waters rose rapidly in the area between the canal and the ridgeline (on an east west axis) and between Brock Creek and the railroad (on a north south axis), ---this area functions like a basin. Unlike river flooding, the canal overflow flood crest lasted a relatively short time and fairly quickly found its way to the outlets which exist at both ends of the basin (one at Lock 5 and the other at Brock Creek). In the area just described the flood surface elevation at some points was several feet high in areas not normally subjected to river flooding, and the level dropped as the water flowed toward Brock Creek ---the river was quite low. The June '96 flood did not reach the ridgeline located between the canal and the river. I personally observed the peak condition along Brock Creek about 8:30 that night from a vantage point on the paved road in front of the flagpole. There was a solid sheet of fast moving water, the flow from both the aqueduct and from the south was toward Brock Creek because starting at College Avenue the land slopes toward Brock Creek. The water surface from the direction of the aqueduct was undulating. My analysis of this flood starts at the Canal Aqueduct Bridge.
CANAL AQUEDUCT BRIDGE
The canal aqueduct bridge has two equally important purposes. It allows boats to pass over the creek, and it allows the creek to pass under the bridge. It is critical that the open cross section under the bridge be properly sized, otherwise the land on the higher side of the canal may become inundated. If the water rises high enough it may overflow into the canal, overflow the towpath, breach the canal, damage the aqueduct, and under the worst scenario destroy the aqueduct.
What actually happened during the flood of June 1996 was that the aqueduct blocked the flow, causing Brock Creek to overflow into the canal. The open area cross section under the aqueduct ---the part that allows Brock Creek to get to the river--- is undersized, probably much less than half of what it should be.
In addition to the undersized cross section under the Bridge Aqueduct, the "velocity of approach" and capacity of the cross section is reduced by a severe creek offset which has developed on the upstream side of the aqueduct. Brock Creek is aimed at the right side of the aqueduct bridge ---in fact the original creek probably followed a path right through the part of the service station nearest the creek. On the downstream side of the Aqueduct Bridge the peak flow is restricted by sharp bends, poor channel alignment, and poor hydraulics between the aqueduct bridge and the Route 32 bridge, impeding the peak flow and causing a backup below the canal bridge, which in turn causes the water to rise even further on the upstream side of the canal bridge. The bridge at river road next to the Yardley Inn will easily take any peak flow, but the potential peak flow spreads out and causes a flood before it gets there.
The canal overflow in June 1996 consisted of surface water originating mostly from the Brock Creek and Silver Creek watersheds. The ratio between the peak flow entering the canal from these two watersheds is not known, but the 3.5 : 1 size of the Brock Creek versus Silver Creek, and the storm pattern on Doppler (as reported by D.E.P. engineers), together with the serious obstructions at the Canal Bridge Aqueduct and backup below the bridge, suggest that the major part of it came from Brock Creek.
Replacing Aqueducts at Taylorsville, Yardley, Point Pleasant, and New HopeExisting Bridge Aqueduct Structure:John Ruddle, general supervisor of canals, reported in 1893...
"On May 3rd last the lower of the 3 spans at Taylorsville Aqueduct broke down, having been in use so long that the timber was rotten; it was repaired by building an entirely new span to replace the one broken down and building a truss on each of the other two spans to strengthen them... This aqueduct is now in fairly good condition, and should last about 3 years more, when it should be replaced with an iron or steel structure. Yardley Aqueduct, which was strengthened during the winter of 1889 and 1890 should be replaced within the next 3 years by an iron structure. During the present winter we are building a new steel aqueduct to replace the wooden one at Point Pleasant and partly rebuilding the aqueduct at New Hope, putting the latter in condition so that it should last from 10 to 15 years longer."
[Ref: "Pennsylvania's Delaware Division Canal" by Albright G. Zimmerman]
Craven's photograph of the Yardley Aqueduct c1910, shows a fairly narrow channel with wood siding, which I'm presently assuming sits inside the same reinforced concrete structure that exists today, and that this underlying concrete structure was constructed sometime after 1893 in lieu of the iron bridge mentioned in John Ruddle's 1893 report. Viewing the bridge from a distance, the concrete abutments are well worn. The steel reinforcing rods are exposed on the broad crested concrete weir on the side along the towpath, where the weir has worn down perhaps 2-3 inches in the center. There was deep erosion behind the wing wall on the downstream side near the sluice gate and the wing wall is badly cracked, and may be seriously damaged. The main force of the June '96 flood flowed over the towpath on the right side in the vicinity of the sluice gate. The massive main structure appears to have easily survived the June 1996 Brock Creek flood. The aqueduct bridge will at some point [to be determined by experts] have to be replaced. When this is done the opening under the bridge will definitely have to be widened, at which time the capacity of the waste weir can easily be greatly increased. The redesign of the aqueduct to accommodate a much larger cross section opening for the creek would provide an ideal opportunity at minimal cost above the basic replacement cost of the large structure, to redesign the entire Brock Creek diversion and reestablish a proper floodway between the canal and the Route 32 bridge.
Aqueduct Bridge Replacement:This project involving both the Canal Aqueduct and the creek channel is straightforward... the creek valley has a large cross section - there is plenty of elevation - the land is open - there will be no houses lost - a culvert is not required- changes to the sanitary sewer should be minimal - the distance between the canal and the river is only 800 feet.Brock Creek Alignment:
In discussing the Brock Creek alignment it is hard to know where to start. The creek between the aqueduct and the river really consists of two separate parts, the normal channel of the creek on one hand, and the larger floodway on the other. The part of the creek upstream from the canal is a straightaway arriving on a sharp angle with the canal, and the opening under the canal has been built on an angle to accommodate it. But the creek is offset. Before the canal was built, the creek looks like it followed a wide curve, but protective work along the creek has pushed it back and created a long straight run behind the gas station aimed directly at the relocated "Toll House" on East Afton where rip rap protecting the Yardley Inn creates an abrupt turn to the left. Following the June '96 storm large size rip rap has been placed to prevent further erosion and undermining of East Afton Avenue. The entire parking lot along the creek is shored up with rip rap along the rear of the parking lot. The creek then heads toward a concrete retaining wall behind the houses on Brown Street where it veers slightly away from the river, and then it takes a final 120-degree turn back toward the Route 32 bridge.
Now lets look at the floodway between the canal and the Route 32 bridge. If you stand behind the fenced "trash pickup area" at the rear corner of the Yardley Inn parking lot, it's a fairly short straight line of sight back to the aqueduct, whereas the creek follows a circuitous route way over on the left. Its difficult to know the original lay of the land, but it appears that a more direct natural floodway is trying to reestablish itself. I think what has happened... starting when the canal was built, man has attempted to contain the entire storm flow inside the creek channel, the result being that any very large storm flows are basically sent in the wrong direction.
Flows equal to June 1996 have undoubtedly occurred many times before the region was settled, although at a much lower frequency of occurrence. There are two river terraces involved here, the old river terrace above the canal and a new lower river terrace starting less than 200 feet below the canal. At both of these levels the creek tends to follow the classic "sine-generated" curve, but the creek valley across the upper terrace (the straightaway behind the gristmill) is naturally deeper than should be expected where the creek crosses the lower terrace, where it should spread out and seek the shortest route to the river. What is needed in the area starting about 200-ft below the aqueduct and extending 600-ft to the Route 32 Bridge, is to restore what nature provided in the first place.
Work between the canal and the Route 32 bridge involves reestablishing the creek channel and the floodway, both on a smooth elongated "S" curve connecting the new wider cross section at the new bridge aqueduct to the already available wide opening under the Route 32 bridge. The sharp bend behind the fenced "trash pickup area" would be eliminated. The creek channel then needs to be shifted away from East Afton, holding the present creek bed elevation, but located inside the floodway to order to achieve maximum peak flow capacity. Parking spaces lost to the floodway can be made up on the reclaimed part of the land and the present planting arrangement extended along East Afton. This would need to be carefully engineered and tentatively approved by the local Planning Commission and the owners of the Yardley Inn property, who are believed to own most of the land involved, at a very early stage of the approval process. See feasibility paragraph below.
Recommendation (for future consideration)...
1) Take advantage of the fact that the Aqueduct Bridge will at some point need to be replaced.
2) Conduct a feasibility study that is "Operations Oriented" (explained as follows):
During the period when the canal was in full operation, canal records show that certain engineers were focused on the canal at the "nuts-and-bolts" level, deeply involved in the operations and maintenance of the canal, working to maximize the use of the existing canal infrastructure, and when necessary making some very technical and innovative recommendations. This "operations engineering" approach is possible where engineers are involved in the day-by-day actual operation of a system over a long term period. Today, the fact that the canal is watered means that it is still a dynamic system, interacting with a constantly changing environment just like it did in the 1800s, and the same approach is needed today.
Start by fully evaluating the original design concept, and the inherent cost advantages of the Brock Creek location (12-ft water surface height directly over the creek and 800-ft distance to the river). Stay flexible. Consider that over the long run, constantly improving storm water management in the local watersheds will reduce the runoff over the next 50-yrs, and over a much longer period may even partially restore the 1800s land surface ---that is one of the possible long term choices
[See "The New Watershed" - Link provided below].
Therefore, maximize the use of the existing system, and avoid conventional drainage practice leading to ever increasing piping and paving. Stay environmentally friendly... consider the possibility of reclaiming the present no-man's land along Brock Creek below the aqueduct, and put the money into a new aqueduct rather than separate new drainage structures, recognizing that due to "watered canal" dynamics and changing environment, it is virtually impossible to keep the canal isolated from its surroundings, or freeze the technology as it was at some particular point in time.
Today, the available excess water discharge capacity at the aqueduct weir is not significant in terms of storms. That's why the canal has to be drained and the weir gate left open at the aqueduct when a storm threatens. A significant part of this constant emptying and refilling can be eliminated if the towpath was raised at 5-7 locations (includes two on the other side), so that a 12-inch minimum capacity is maintained above the weir. This may involve as much as perhaps 1,000 total linear feet of the towpath. The project needs to be surveyed, engineered, and specifications prepared to insure the same locations don't easily fail again. Work needed includes surveying levels around the perimeter including the low spots on the opposite side of the canal, and the ground outside the towpath. Much more extensive modification should be considered at the time the aqueduct is replaced so that even further advantage might be gained.
Feasibility StudySTEP 1: Determine the maximum size of excess water diversion that can be created at Brock Creek.
STEP 2: Depending on the findings in Step 1, determine the remaining excess water diversion capacity needed, and how much can be created at Lock 5. There is an existing towpath overflow including: (1) - Overflow at partial "spillway" across towpath on a 150-ft front just up from the entrance to Lock 5; (2) - Existing man-made basin 6-ft deep, normally dry, created by the
interposition of the railroad in 1875, (3) - Basin spillway 3-ft above bottom of basin and 3-ft above the downstream water surface at lock, (4) - Slow release pipe under the railroad, (5) - Main release across open ground between the railroad and the canal. Further down past the walkway into the park, there is what appears to be an abandoned channel. This may be remnants of Silver Creek which is said to have been flowing that way [particularly during storms] during a 50-yr period before the canal was constructed. (6) - A normally dry ditch below the railroad connects the pipe under the railroad to the dry channel just mentioned. (7) a natural swamp starting somewhere down near the lower end of Macclesfield Park, (8) - Potato Creek drains the entire area to the Delaware River. The amount of the overflow which occurred at Lock 5 on June 19, 1996 should be taken into account in the water balance.
Water Balance:
Silver Creek + Lock 6 (Bypass + Leaks) + Local Drainage = Lock 5 (Bypass + Leaks) + Lock 5 (Overflow) + Flow Diversion at Aqueduct Bridge
Step 3: Canal channel hydraulics factor: During most storms the water level in the canal is fairly steady or slowly rising. It is only under severe storm conditions that the flow velocity increases enough so that the level of the water may vary significantly as it moves along in the canal. How the velocity is maintained, the height of the towpath, splitting the flow in two directions, and whether the towpath is on solid ground or on the top of a berm are all factors involved. The hydraulic profile along the canal channel should be established for a 10-yr, 25-yr, 50-yr, 100-yr flood.
Step 4: Towpath
Field survey and design towpath modifications. Bear in mind dimensions of the various parts of the canal as they exist today, may not be the same as the original dimensions.
Step 4: Compare alternate plans
Compare the cost of the proposed project based on the aqueduct replacement as outlined here, to any alternate 100-yr solution based on constructing an entirely new and separate storm drainage system. In making the comparison, take into account that the Bridge Aqueduct will eventually need to be replaced anyway, and/or that one approach gives you a new aqueduct bridge, depending on how you look at it.