Loading...
HomeMy WebLinkAboutAPA2687·fNGINEER MANUAL [M]&OO~& c ~®£®©@ Susitna Joint Venture Document Number EM 1110·2 .. 1612 15 ClC:T 1982 ',:</ .f ..,. • ' n " . • . -Please Return To !!_ DOCUMENTCONTRO_L___a~-------------·----;, . -., . .. ENGINEERif#~G AND DESIGN ICE ENGINEERING / .. ' . . . . . . .. ,. • o ' ,. I ... ~ • o ' .. . . • . . 6. ,.. ~ pG " ) r -~ .. "' • DEPARTMENT OF THE ARMY r'n~~~~ n~ ~~~~"~~~P~~ E F FR f g H h ht hj lt L m n p p Q R Rn s s T t u v z a e 'Y e ). -----• = ---• -----• ----= - ---• • p - t,a • 1' (1 lbf -lbm • Young's l'DOdulus Fahrenheit Froude number Friction factor or force Acceleration of gravity Depth Ice thickness Heat transfer coefficient Thickne~s of ice jam Empirical coefficient Length Meters Roughness coefficient (Manning's "n"} Pressure Wetted perimeter Discharge Hydraulic radius Reynold's number Hydraulic slope Seconds Temperature Time Wind spcad Velocity or volume Accumulated freezing degree-dayP Ang'le Angle Specific weight Porosity Latent heat Density Strength Shear stress Stress pounds force • (lbm) x (32.2 ft/s 2 ) pounds mass 1-9 EM 1110-2-1612 15 Oct 82 CHAPTER 2 ICE FORMATION AND CHARAGTERISTICS Section I. Ice Types EM 1110-2-16)2 15 Oct 82 2-1. Introduction. Ice occurs in a number of forms which are not fa~iliar to everyone. The following details about the material should be revie•·ed before any effort is made toward the study of ice or its effects. a. Ice grows in hexagonal crystals. There are three "a" axes of s~try in what is called the basal plane and one "c" axis perpendicular to the basal plane. An ice crystal looks like a pencil, with the "c" axis as the lead. When a calm body of water starts to freeze, the crystals nucleate with axes oriented randomly, but since growth is easiest along the. "a" axes, crystals ~th horizontal "c" axes grow faster and gradually, with increasing ice depth, predominate. However, ice crystals with vertical ·c· axes are also common. Crystal growth is a natural refining process that rejects impurities, such as the salts in sea water. Most impurities stay in the unfrozen water, but some are trapped between the individual cry~­ tals. Because of the trapped impurities-, melting begins at the crystal boundaries and a phenomenon call.ed candle ice oft~n develops. In candle iet!, ~tttl'\it~n~rable singl~ ~cystals are uo ltJnger frozen, togethetJ ~.J~ fltlu~fi are leaning on each other for suprort. A small wave can collapse ene entire mass like a +~~~ Qf ~Q~.nOf#ll! b. Much of the ice on a lake or river takes a different form called snow ice. This material ia granul;a.r, opaque, and white., There are no large crystals so melting is not a~ spectacular a process; there is no candling. Snow ice is formed when a snow cover is saturated by rain or by submersion in a lake. Compared to crystalline ice, snow ice has more grain boundaries which makes the jLce weaker, and it is isotropic. Snow ice and massive, consolidated frazil slush look much the same. Frazil ice, which will be covered in more detail later, results from small particles of ice forming in supercooled, turbulent water-. 2-2. Sea Ice. Sea ice is quite different from freshwater ice. Whi1~! it is growing, plate-like crystals form a mushy layer about 1 inch deep that precedes the ice front downwards. Some seawater is trapped between these platelets, leading to many liquid inclusions or brine poCkets in the ice. As the ice gets colder, water freezes to the walls of these inclusions and the brine becomes more concentrated; as the ice warms, the walls of the inclusion melt and dilute the brine~ Since these po~kets are part of the total ice cross section, the strength of sea ice varies more dramatically with temperatur·e than that of freshwater ice. In passing it should be mentioned that seawater (3 .5 percent salt~,) freezes at 29°F and is most dense at 29°F, unlike fresh water which is most dense at 39°F. As we all know, water expands when it freezes, unlike most other materials. However, once it is ice, it shrinks as it gets colder like any other material. The bottom of an ice cover is always at the freezing point since it is in contact with water. Changes in air temperature~ therefore, result in 2-1 EM 1110-2-1612 15 Oct 82 bending and thermal cracks. The effect of these thermal cracks on the engineering properties of ice is not well known. 2-3. Frazil Ice. Frazil ice occurs in two distinctly different atates: active and passive. The acti!le state occurs in aupercooled water as aull ice nuclei form and grow into discoid& of limited size; Typical supercooling in natural bodies of water is within several hundredths of a degree below the freezing point. The limit of the discoids' size is dependent upon growing conditions but usually does not exceed about an inch in diameter by a few tenths of an inch in thickness. The active frazil is highly adhesive. As the water temperature riseB to the freezing point, discoid growth is curtailed, and the passive state occura. The diacoida then lose their adhesiveness. ~· Prazil ice production requires both supercooling and nucleation. An ice cover inhibits supercooling by insulating the water from radiative and convective heat loss. Any supercooled water passing under an ice cover will freeze to the cover itself instead of forming frazil. ~ome investi- gators believe that frazil pr~duction takes place in a thin surface layer of supercooled water. An ice cover precludes the existence of such a layer. Surface turbulence, caused by wind, flow patterns, or other factors, will prevent an ice cover from forming and 1ncre&8e frazil pro- duction ... b. Any re.asonably fast flowing (2.0 feet per s~er:ond (fpe) or .,re) stream can be considered a frazil producer when air ~c.··:,~rAtures are below ~beu~ 20~F. Lakes ana reser1oirs, where waves provide eurface turbulence and mixing, also produce frazil. c. In addition to boundary growth, aetive frazil particles may agglo- merate. Active frazil will remain in suspension ·until the buoyant effect of increasing size overcomes the submerging forces of gravity and turbu- lence. If the frazil surfaces in below freezing weather on open water, 1~ will continue to agglomerate. Field observations indicate that as the stream velocity slows and allows the frazil to surface, the surface turbu- lence decreases and allows formation of an ice cover. Frazil may then accumulate under the existing ice cover to form an obstruction called e hanging dam .• 2-4. Hanging Dams. A hanging dam is a downward projection under an ice cover formed from active frazil. Brash ice brought 4ownstrea~ during the formation of the dam sometimes lodges in the frazil. Frazil particles will continue to accumulate into a rigid structure until some force overcomes the bonding process (Figure 2-1). The bonding process is usually overcome when increasing water velocity ~due to the decre~sing cross-sectional area of the stream) is sufficient to c~rTy frazil particles under the dam faster than they can attach to the body of the dam. 2-2 Figure 2-1. Formation of a hanging dam. EM 1110-2-1612 15 Oct 82 a. The growth of a dam may cause dramatic scouring in the river bed as bottom sediment is eroded by the increased water velocities under the dam. As with any restriction to flow, hanging dams tend to cause backup and reduce the watP.r level downstream. In Alaska these dams have broken loose during freezeup and caused some. f!9Qd!ng: During b!'@3KUP; jamming can occur for miles behind a dam, and ifi e~e~ijijtve head UP.$tream aay be required to break through the dam site. Failure of the ice at the sit~ ia primarily a function of the watershed runoff. 1f ~uuff i8 s~adual and consistent, the dam may "rot" in place. On the other hand, if the diurnal melt is significant, the chance of a head buildup behind the dam in- creases. If the head is sufficient, a dam may fail dramatically, causing large ice runs and associated flooding. The h~ad buildup may be required simply t:o dislodge and move the large mass of ice at the dam site. Shear failure often occurs on the sides or bottom because of the different chan~ nel dimemsions at the site. If the runoff is such that the frazil body of the dam is allowed to gradually rot, the bottom shearing and head buildup are reduced or eliminated. Under heavy spring runoff conditions, it baa been reported that dam breaks have run 25 miles of river. b. Dam formation has been unpfficially enhanced at particular sites in order to control downstream levels. This use of dams is urg.:.nally effective because of the diversi.ty of freezeup patterns in most rivera and streams. c. Detection of hanging dams was previously limited to the recognition of characteristic sites~ Often dam sites are characterized by a hummocked surface, similar in appearance to that of an ice jaa during breakup. This. bas been determined to be an effect of the buoyancy of a rapidly growing dam under a reasonably thin ice cover. This characteriat!e will normally be somewhat obscured by the season's snow cover since the surface is formed in early winter. These sites are typically deeper, broader areas in the channel where a significant reduction in flow rate 11 observed. They must be preceded by a frazil-producing area. Manual sounding has been the only method available for widespread asppiog of ice 2-3 EM 1110-2-1612 15 Oct 82 accumulation. During the past few years a radar ice-thickness profilometer h~s been tested by CRREL under various circumstances to determine its applicability for frazil ice detection and ice thickness measurements in fresh water, The ·results of the field work were encouraging. Such an instrument used in a remote sensing role was shown to be advantageous for investigating a large or relatively inaccessible area and for searchir.~ out areas of ice accumulation. The system evaluated is commercially available (being used for near surface geophysics), but is still in the developmental stage for ice applications. 2-5. Oth~r Frazil Problems. In its actlve state, frazil will adhere to nearly all submerged objects, clogging channels, trash racks at hydro- electric generating stations, municipal water intakes, and other submerged structures. a. Submerged objects can be protected from active fratil. Techniques include the use of low thermal conductivity materials and ·sacrificial" coatings such as a special silicon grease. The obvious way to eliminate ice adhesion is to heat the submerged object. In some cases, the resources are available within a plant. For example, the heated discharge of an atomic power plant may be circulated thYou.gh the trash racks of its in- take. In other cases, when electrical energy is required, costs may be substantial; so a feedback system which !DOnitors the temperature of the object should be used to minimize the heating expense. b. Even if the frazil is prevented from adhering to an object, its sheer-bulk {Uln often clog intakes. In extreme cases entire generating stations have been forced to shut down for lack of flow. Northeru dis- tricts have reported lock operation and maintenance problems because of frazil a.ccumulation. Also, frazil accumulation in certain navigable water- ways has become so extreme that river tow boats cannot maneuver. 2•6" Engineering Cha'racteristics of Frazil Ice Accumulation.!• Unless frazil ice has accumulated because of high supercooling and/or high hy- draulic forces (such as is the case with anchor ice in a rapids reach), its initial structure is quite weak. Apparently driven by heat transfer from the surface, the weak accumulation ages and attains significant strength. This aging may occur over a few weeks. The in situ forces on the accumu- lation will also affect the characteristics through deformation and con- solidation. Field measurements of the engineering characteristics of such accumulations prov.ide the following parameter ranges: shear strength • 1-8 psi (pounds per square inch), bearing capacity~ 15-45 psi, density • 4D-70 percent ice by volume, and permeability comparable to a coarse sand or a fine gravel. Section II. Ice Growth 2-7. Calculating the Quantity of Frazil Ice Generated!_!! a Water Body. An accurate assessment of the potential problems caused by frazil requires 2-4 EM 1110-2-1612 .15 Oct 82 a numerical estimate of the amount of frazil generated in a body of water. While me.thods are still cr\lde, a comparative estimate can be obt~ined from the relation where Vf gs A Pi Af llT t v -f g A ~T s • • volum1e of frazil ice pr:-oduced per second, ft 3 •' beat transfer coefficient, BTU/ (a oft 2 • ~F) • open water area producing frazil, ft 2 .a maes density of ice • 57.2 lbm/ft !~ • laten.t heat • 144 BTU/lba (1) • average temperature below 32°F ~~ring period of interest, •r • time, seconds(s) The volume of frazil ice produced is not the volume that it wtll occupy when it accumulates. Early field ~ot·k has shown the porosity of frazil accumulations Sf to be between 0.4 and 0.6. The aecumulated volume, Vfd, will be approxi1nately twice Vf• The portion of the total flow lost to ice generatic>n will be Pi/ Pw Vf, in cubic feet per second (cfs), where Pw is the density of water (62,4 lba/ft 3 )! .!.• 'rhe heat trstnsfer coefficient is a function of several paraaeters. For esti.uting ice generation over several weeks! field experience has shown that a value of 9.8 x 10~4 to 1~23 x 10- BTU/(s•ft2•0F) is reasonable far east North American rivers and streams. If extreme condit~Qgs e~ist (i~ee. sustained high winds) a site specific value of this parameter must be used. b. The area of open water producing frazil is probably the most error=ridden term in the computation of frazil production. There is an intricate relationshi:p between the amount of ice generated and the extent of the open areas pro•!ucing ice. The error associated with iaproper estiaates of this relationship is minimal in broad, uninterrupted open reaches. It becomes significant when meandering open reaches are edged with sheet and frazil accumulations and are bridged with sheet ice and a now. £• The temperature may be varied over small periods of tiae if short-term variations in ice generation are desired. Over several weeks, however, daily fluctuations tend to cancel each other, and the use of monthly temperature everages provides reliable data. 2-8. Exa11ple Problem. A reservoir fed by .a 200 to 250 cfs mountain stream can be constructed in a nu111ber of ways. The evaluation needed for wintertime operation requires an estimate of frazil generation. The reach of interest is not extreme and may lose 1.0 x l0-3 BTU/(a~ft2•°F). Local 2-5 EM 1110-2-1612 15 Oct 82 temperature records show that the average temperature for the .anth of January is 26°F. The model describes the •tream 88 being 3 feet wide and open along a 11eandering course for 15 11iles. a. A normalized value may be calculated that estimates the production of ic; in terms of the volume of fra~il per unit surface area in square feet per day per °F. By changing the value of variables, one may also establish the sensitivity of the model's parameters. b. -For the present problem the normalized frazil ice generated VfN is V • gsA AT (2) fN Pi ).f • (l.Oxlo-3 BTU/[s•ft 2 •°F])(8.64x10 4 s/day)(l ft 2 )(1.F) ("s7.2 lbm/ft 3) (144 BTU/lbm) · • 1.05 x 10•2 ft 3/(day•ft2•°F) i.e., the volu~e of frazil ice generated per square foot of open surface area, per day, per °F below 32°F is 1.05 x 10-2 ft3. For the particular proble~, where AT • 6°F v f -v f X AT X At X AA N At • 31 days 6A • 3 ft x 15 miles x 5280 ft/mile • 2.38 x 10 5 ft 2 The solution is V£ m (1.05 X 10-2 ft 3 /[day•°F~ft 2 ]) (6°F) (31 days) (3) To find the volume, Vf , occupied by this quantity of frazil ice when it D deposits in the stream, divide by the frazil porosity, 8f• Recall that the range of frazil porosity is 0~4 to 0.6. On the average then, the deposited frazil volume is v£ • v£ I af D • 4,65 x 10 5 fe 3 /o.s • 9.3 X 10 5 ft3 (4) £• Figure 2-2 provides a family of curves b~sed on the amount of open water area during the period of frazil ice generation. The figure plots 2-6 EM 1110-2-1612 15 Oct 82 the volume occupied by the accumulated frazil ice as a function of the freezing degr·ee-days. The family of curves should not be used with smaller open areas (less than 10 4 square feet). Frazil generation in such small areas is extremely sensitive to the microclimate and cannot be predicte~ from average parameter values. d. As an alternative to calculations, one may solve a problem using the previous example. Locate the freezing degree-days (~T x t • 6°F x 31 days • 186°F days) on the abscissa of Figure 2-2. The ordinate value of the intersection of this line with the corresponding open water curve (2 .. 38 x 10 5 square feet) is Vfn~ Note th&t the open water curve is interpolated logarithmically. ~ 10 1 d I r;i) "' .!: e ::r .. .. 2 10 7 "'--... 0 ... 1.1.. >-&1 ~ !! 10 6 ~ ::r ~ u 0 il E ::r ~ I 10 5 > ... o Figure 2-2. Frazil ice accumulation as a function of degree-days. 2~9. Estimation of lee Formation Dates. A comparison of the long-ferm freezing degree-days• curve (Table 2-1) and the average ice formation date yields estimates of when ice ~~11 form and how long it will remain. The calculations require long-term (30 years) air temperature normals and dates of ice formation in previous years. *A freezing degree-day is a measure of this departure of the mean daily temperature from a given base (32°F in this case). 2-7 !M 1110-2-1612 15 Oct 82 TABLE 2-1 · Example of freez:f;;:g degree-day calculations. Freezing Day Temperature (OF) degree-day E degree-day Min. Max. Avg~ -~ 1 -10 0 -5 37 37 2 -15 -s ··10 42 7ql 3 -s +11 +3 29 lOfl 4 0 +10 +5 27 135i 5 -10 0 -s 37 172 a. These values are plotted thro"gh the winter. Each day on the curve represents an accumulated valu~ of freezing. Figure 2-3 shows an example of accumulated ~reezing degree-days at two sites during the winter of 1976-77 as compared to the normal, along with pertinent notations of da.tes of ice formation. J?.• Lake Champlain! Vermont, for example, is a site that provides sufficient .ice and temperature data. Burlington aitportt n~a~ the lake, has ~xcellent long-term temper&ture records and a meteorological stat~on has been installed on the lake shore for compar.ison. Computed freezing degree-day records (Table 2-2) were correlated with data on the ferry's closing and opening dates (because of ice) for 19 year;@ (3~" r base) ( 0° c bose) • • ~ Cl • c Cl IC N "' ~ ~ LL. • > -Cl ::::t e ::::t (.) 400 (i) F1rst Ice (Z) Grand Isle Freeze· Over @ ® ® -Bul'hnoton --Grtlnd Isle ·----Normal Jan 1977 Figure 2-3. Accumulated freezing degree-days at two sites: Winter 1976-,77. 2-6 EM 1110-2-1612 15 Oct 82 c. Table 2-2 shows that on the average, a permanent ice cover forms on Lake Champlain at Grand Isle on 15 January ± ~3 days, with normal ice-out occurring on 5 April ± 13 days. The lake n.o:rmally closes for navigation after 615 freezing degree-days ± 157 degree-days. Tahle 2-2 also shows that the lake is normally closed for 82 days. To estimate when ice will occur during the current season, we start in Nove1nber and select either the normal freezing curve or the annual curve (Jf the 19 available} that best .fits the current freezing conditions; we then follow this accumulated plot to the ice formation date. d. As better air temperature forecasting methods (for 30-6Q days} become available, the above method will predict ice formation dates with greater accuracy. TABLE 2-2 Historical ice data+ for Lake Champlain. Date closed (freeze-over Date open ~t ferrtl (f!r;x start) 1/8/60 1/10/61 12/31/61 12/30/62 12/31/63 1/15/65 1/26/66 2/6/61 1/8/68 1/9/69 1/5/70 1/17/71 1/29/72 1/9/73 2/7/74 2/4/75 1/13/76 12/28/76 1/16/78 Average: Jan 15 ( ±13 days) 4/2/60 4/15/61 4/14/62 4/i5i63 3/26/64 4/15/65 3/25/66 4/5/67 4/3/68 4/8/69 4/19/70 4/26/71 4/26/72 3/16/73 3/26/74 3/29/75 3/31/76 3/21/77* 4/8/78* Apr 6 (±13 days) Number days lake closed 85 89 104 107 87 9! 59 59 87 90 105 101 89 67 48 54 79 83 82 82 Number freezing degree-days to ice cover . Date lake completely frozen Number freezing degree-days to close of entire lake 468 594 324 396 522 540 558 Lake didn't close 720 612 765 666 990 729 558 864 666 666 450 576 1/27/61 1008 2/16/62 1152 2i8/63 Lake didn't Lake didn't 2/7/66 2/13/67 2/16/68 3/2/69 !/21/70 2/2/71 2/10/72 2/21/73 2/15/74 2/21/75 Lake didn't 1/18/77 2/13/78 i044 close close 864 909 1386 1404 1174 1368 918 1134 1026 936 close 882 1080 615 Feb. 11 1085 (±18 days) ( ±87 deg. ( ±99 deg. days) days) * Ferry navigated in ice cover these two winters. + All temperature data are from the Burlington, Vermont, airport. 2-9 EM 1110-2-1612 15 Oct 82 2-10. Prediction of Ice Growth. The date of ice formation and the initial ice thickness at freeze-over are needed to estimate ice growth for the case cited in this summary. a. Figure 2-4 is a plot of actual measured ice thickness against computed growth for Shelburne Point, Vermontf on Lake Champlain. The following simplified Stefan equation. determined the computed growth of the ice cover: h(t) -kZ 1 ' 2 (5) where h is the thickness of the ice in incbes at time t, Z is the total number of accumulated Fahrenheit freezing degree-days since time of perma- nent freeze-over, and k is an empirical coefficient. Figure 2-4 shows an estimated 1ce growth curve starting on 13 January (first permanent ice) based on observed temperatures and an empirical coefficient of 0.6. Ttie two curves in Figure 2-4 coincide fairly accurately. The predicted curve reaches a maximum thickness of 13.4 inches on 15 February, while the actual maximum thickness was 13.8 inches on 19 February. The two curves are compatible until 8 March when strong northerly winds broke up the ice at Shelburne Point. (em) 50~~~~-r~r-~~,--r~-r~-,--r-r- • Ice Growth 1 • i 1 ········Computed h(t)= kZ~ k=O.S._Oay-1-F-2 -Actual ---In/out S1!uation (Flowing Slush Ice) .•• ... )'\ . . . : _.,., ... j.P'' •. . · '• ·. ~ . . , .. · .. ; . Figure 2-4. Actual measured ice thickness vs. computed growth at Shelburne Point, Vermont: Winter 1975-76. k· The following year when the measurement site was moved to the Grand Isle, Vermont, location, the empirical coefficient for best fit of the ice growth curve was calculated at 0.3. Caution should be used in . stating an empirical coefficient for a large body of water such as Lake Champlain. The empirical coefficient changes with wind exposure and geographical lo~ation of measurements on such a lake. 2-11. Thermal Strains in Ice. At 32°F and atmospheric pressure, the specific gravity of water is 1.00 and the specific gravity of pure ice is 2-10 EM 1110-2-1612 15 Oct 82 91 7; this means that freezing gives a volumetric strain of -8 percent ana-O~at~ing gives a volumetric strain of +9 p~ercent. As ice is cooled at t stant pressure, it contracts. The expansion coefficient varies with con temperature, but in ordinary ice engineering this variation can usually be 1gnored. a. For a single crystal of pure ice, the coefficient of linear expansion varies slightly with crystallographic direction, but at temperatures near 32°F the difference is only about 2 percent_ (greater parallel to the c-axis than perpendicular to the c-axis) and can be ignored for most practical purposes. At temperatures between 32°F and -40°F, we can take the coefficient of linear expansion as equal to 2.8 x 101-5/0 1 (±4 percent) at 32°F for freshwater polycrystalline ice, noting that it decreases by approximately 10 percent as temperature drops from 32° to -4o•F. b. In sea ice, or other ice containlng significant amounts of dissolved impurities, the situation is very different. At temperatures near the melting point, saline ice consists of solid ice plus liquid inclusions which change in volume and salinity as the temperature changes. Although the ice crystals have a positive expansion coefficient, phase changes in the brine cells creat2 freezing strains iri a negative sense, i.eg the volume increases as temperatu~e drops. Above about 23°F, the r~eezing and t~~wing of brine ifieluiions is usually thought to be the dominant effect, implying that sea ice contracts as temperature increases in this range. At very low temperatures, when all the brine is frozen and solid salts are precipitated, the behavior is almost identical to that of -pure toe. 2-11 CHAPTER3 ICE JAMS 'EM 1110-2-1612 l5 Oct 82 3-1. Introduction. When the ice goes out of the rivers it often jams and causes flooding of fields and homes. The ice floating upstream from a jam can destroy housest bury roads, and collect on fields, delaying spring plowing and planting. Figure 3-1 shows how massive this ice can be. 3-2. Diacusuion. Th~re are two processes which either alone or in concert are responsible for breakup. lc:a strength grs1dually deteriorates in the spring when higher sun angles and higher air 1:emperatures melt snow from the ice surface, forming a water layer. This water layer absorbs more solar radiation, causing subsequent melt along the crjstal boundaries. If not disturbed by other factors, the ice will melt in place. a. In rivers, the current flow ~neath the ice i~ a second factor. In fact, water flow is the sole cause of the midwinter breakups that can lead to the most destructive ice jams;. Any inct'ease in water flow down the ri,1er will raise the ice level and break it loose from the shore. If the rtver discharge stays high because of rain or snow melt on the uppeJC sec- tions of the watershed, the higher flow will move th~ ice downstream. As te moves, the ice breaks up; the size of the pieces depends on the distance they move and the degree to which the ice strenr;th has deteriorated. A.s might be expected, the ice in those reaches with steeper sl•Dpes and hi~h~~ curreat velocities will go out first. When tne moving 1~~ hits the fixeq ice in a ,low, flat reach, t~ may break up th~ stat~onary tc~ and c&r~/ it along, or form a jam~ Ie~ j~ma occur in two basic forms, the dry jam and tb! simple jam. Th~y are essentially identical except that in a dry jam the ice is grounded, restricting water flow to a greater degree than a simple jam. b. P~edicting the time or even the probability of an ice jam occurring is still uncertain. However, there are a number of typical locations where a jam will form. As mentioned earlier, any section of a river where the slope decreases is a possible location. During freezeup the slower moving reaches freeze first, and so will have a thicker ice cover come breakup. Another possible location might be a constriction in the channel, either natural, such as at a bend or at islands, or at man-made features such as bridge abutments and midstream piers. A third typical location is a shallow reach where the ice ~an freeze to bottom bars or boulders and will not be lifted and moved by the increased water flow.· c. Once the ice is stopped at any location the jam thickens rapidly, primarily by ice blocks turning under existing surface ice. The net result is a very rapid constriction of the channel and subsequent baCking up of the upstream flow. d. Flooding from an ice jam occurs very quickly. The situation is not like a normal &~en water flood when the channel is not large enough for the flow. Instead the channel is completely blocked. Normal backwater calculations based upon stage level recorders are meaningless. Suddenly 3-1 EM 1110-2-1612 15 Oct 82 there is a new dam in the river, albeit a leaky and temporary one, which ls creating a lake and which has no convenient spillway. The beet time to try to .ave a jam is while the water pressure behind it is still high and the flow rates are adequate to carry the ice downstream. If the jam occurs in midwinter and a cold spell reduces the flow before the jam moves on, it can settle an the bottom and remain for the rest of the winter, creating a potential hazard. During the t·est of the winter a new ice cover can foru1 upstream, and when the spring breakup comes the new ice cover will be stopped at the old jam; flooding is almost a certainty. The ne~d to fre!~ or remove some 1,-:e jams is thus obvious; both in caseo when f!.ooding is actually present and in cases when the potential for subsequent flooding exists. - . t: -~· ~~" , . ... Figure 3-1. Stranded frazil ice at Cattaraugus Creek near Buffalo, N.Y. 3-3. Methods of Ice Jam R.emoval. There are four different methods for removing ice jams or alleviating ice jam problems, and each has its advan- tages and drawbacks. These are mechanical removal, dusting, blasting, a~d the use of icebreaking ships. It is important to remember that ice loosened in a stream may jam elsewhere. A decision must be made. Is it best to move the jam and talte the possible financial responsibility for downstream damage, or to accept the potential damages cay.sed by the jam HS is? Once the decision has been made to try to remove the jam, which approach will be the most effective? a. Mechanical Removal. Removing the jam mechanically, for lack of a better t~rm., means simply taking the ice out of the atream bed and placing it elsewhere. This, of course, eliminates any downstream problems but it is neither cheap nor fast. In February 1978 it cost approximately $11 .501) 3-2 EM 1110-2-1612 15 Oct 82 tc .ake 1 ~.6®-f~~ ~h~~~l ~;h one Caterpillar 235 backhoe. The approach ia further liaited to dry j.-e in relatively shallow atrea... In other warda, thi1 approach is uaed generally for ~dvinter jama on ... 11 1 tre... after the flooding ha& receded. the idea is to create a small ehanrtel within the jam by using mechanical equip.ent such as bulldozers, backhoes, or draglines. When the ice blocks are s.all and thin, .echanical clearing does not present too great a problem. When the bloeka are around 10 x 10 x 2 feet or larger, small equipment ia generally inadequate. Each site is different, so that equipment and methods used are up to the operator. Be aust be aware of the problems of power linea, poor bottoa, and access. An i~~~~ediate problea is disposing of the iee. Usually it can be pu8hed to each side, leaving a channel about one-third the noraal river width. In reaches where the channel has been severely restricted b7 s&n-made works, it aay be necessary to reaove all the iee. !.• Dusting. A &econd aethod for alleviating ice jaa proble• is the use of dust.-By dust we •an any clark aubstanc:e that can be apread on the • ice in a thin layer to ab•orb solar radiation and thereby hasten the deterioration process. This method is uaed pri .. rily to alleviate poasible ja conditions before the fact. The rough surface of an act~al j• cre8ltea so .. n; shadows that the dust is not effective. lor example, Moor and Watson describe a reach of the Yukon River down8treaa of Galena, Alaska, which has regularly caused. ice jama. Dusting this n:aeh each Sl»ring two to three weeks before breakup weakens the ice sufficiently that no jams have occurred there since the practice started. Ideally, the ~~8t should be applied aa early as possible but after the last snowfall. In general, any reach with an ice cover that regularly stops .the ice run and causes a jam could be weakened in thi3 manner. (1) Dueting involves spra-~ing (as evenly as possible) 8 dust or sand layer and letting the sun provide the energy. Thus, time is involved aa well as the higher sun angles in the late spring and good luck in avoiding snow ator.ma which would cover the dust. Agricultural aircraft gene~ally apply the dust, which keeps coats f8irly low. Moor and Watson 7 give a coat of 34.9 cents (1970 dollars) per lineal foot (100 feet widG) in a remote section of Alaska. The particle aize can vary·, depending on what is avail- able. Moor and Watson 7 quote 0.3 pounds per square yard for sand and 0.35 pounds per square yard for fly ••h. V.I. 51notin3 gives similar rates: for 0.04-inch dia.eter dust he suggeats 0.18 paands per square yard and for 0.2 .... inch duat, 0.92 pounds per square yard. (2) A logical offshoot of dusting is to puwp water and bottom aateriala onto the ice eurfacen This is liaited to streams with silt or ssnd bottoms and, according to Moor and Watson 7, is ten times aore expensive than aerial ~~t!~g. However, the approech does ha~e application where the stream is too narrow or sinuous for aerial work, or where envirou.ental considerations preclude adding .. terial to the streaa. ~· Bla•tiy. The third method 1a blasting the jam. For immediate flood relief thie is probably the aost effective. The priaary purpose of the blasting ia to loosen the ice. However, enough flow aust be coaing 3-3 1M 1110-2-1612 15 Oct 82 through the ju to float the iee downatreaa. Thus, a prerequisite to blaating ia an ice-free reach dovnatreaa where the lee can go, either all the way down the ri var, or to a apot where it •.rill llOt ju (or, if it does, where the j .. will not cause any appreciable daaage). Unfortunately, j ... hav.e been bl&{tted without regard to downatream problea. Succeacful blaa·tiog takea time and careful planning. { 1) The ideal tie to blaat a jaa is just after it haa foneci. In actuality, a jam is never blastad this quiCkly beea~se a blaatias crew and governaental approval caanot be .abilized until the jaa 1a well fo~d and flooding ~ begun. If the flow bas dropped becauae of cold weather or has •oved into another channel ao that after a blaat there will aot '-enough water to carry the looaened ice downatreaa~ the blaatlaa abould be can- celed. (2) If the deci.eion hu been ude to bla•t, there are a auabar of pl'."o- cedures learned froa experience that can lead to a ch~apar &ad .ore 5Ue- cessful job. Each charge, if placed under the ice, vi:Ll blow a crater or circle in the ice with a clia.ter that is related to tbe wipt of tbe explosive. J'igure 3-2 gives this relationship. A bancly cbarp aize for 110st jobs is around 40 pounds, which gives a diaMter of clc•e to 40 feet. Experience bas ahown that two 110re or less paralal ratra of charsea, aet close en()I,Jgh eo the craters intersect, give th& beat result. If it is poasible to locate the thalweg or deepest part of tbe river the blasting line. should be along it. This creates an open Channel with good flow dept.hs that is vide enough to preclude aost aeconciary j...tq. Tbe charges must be placed in the water below the ic~ cover. This 1a eztre.ely t.por- -"' ... -. :2 --.11: .!' 6 • • & ~ J u -a c • .a :I u -.11: 0 i ' ... J u Figure 3-2. Kelationahip of explosive weight to crater hole (good for ice 1 to 10 feet thiclt) • EM 1110-2-1612 15 Oct 82 -tant since the driving force is apparently the large gas bubble resulting from the blast, and not the shoCk waves. The charges must be weighted to sink but also roped to the surface to keep them from being carried downstream by the current. (3) Any kind of explosives can be used for this work; however, from experience ANFO is preferred. {The WTiter has no experience with slurries that aay work as well.) ANFO is 5 mixture of ammonium nitrate fertilizer and fuel oil. The best ratio is 6 percent ~ weight oil with prilled (in pellet fora) nit~ate. This ratio works out as 1 gallon of oil per 100 pounds of fertilizer. The mixture must be detonated with a strong booster such as a stick of dyne~te, TNT, or the special booster charges sold by the powder co8panies. Like many other explosives ANFO must be kept relativeli dry, so placing the mixture in a plastic bag whiCh can also hold a brick, sand, or whatever weight is necessary to sink the charge is recommended. ANPO is relatively cheap and it will dissolve with ti.e if a aisfire takes place~ and not leave large, live charges on the river bottom. As a guide, it is preferable to use Primacord for all downhole and hookup lines. This is then set off with one electric cap vhich is taped to the Priaacord at the last moment when the rest of the party is off the ice. (4) Blasting is not a quiCk, easy solution. It requires some planning to locate and acqui.re the explosive, the equipment to aake holes to place the charges, acd manpower. At all times when the crew is working on the jam, a lookout should be on duty some thousand feet upstream to sound the alarm if the jam lets go by itself. At least two .en are required to drill holes, and depending on the roughness of the surface, at least four aore to carry the charges to the holes. Add a blaster, a supervisor, and two men to load the charges and you have a crew of 11 people. With good luck this crew can blast two rows of charges along about a half mile of river per day, poasibly more when a routine has been established. (5) A formal safety plan covering all operations is necessary. It should comply with both local and F~deral regulations. Such matters as person in charge, communication, transportation, warning personnel~ etc. should be fully covered. ~· Icebreakers. The fourth method of removing jams in only usable in a few rivers. When the channel depth is sufficient and the ships avail- able, icebreakers ar·e certainly the easiest, safest and cheapest way to break up a j aa. This operation is carried out ~ the captains, who are responsible for the safety of their ships, so little more needs to be said rega~ding safe operations. If two ships are available, they work best in echelon (staggered one behind and to the side of the other), starting from the downstream end of the jam. The following ship bas to be careful to ensure an equal width channel. If it crosses the path of the leader, the resulting narrow section will inevitably cause a jam and the downstream channel will no longer keep itself clear. Occasionally, if circumstances permit, an icebreaker can work in conjunction with bleating. The propeller 3-5 ' EM 1110-2-1612 15 Oct 82 wash and vcve action of the ship will clear the ice looaened by the bl5atiag faster, and the ship will offer a factor of safety for the people on the ic~. A combined operation like this will require extra cooperation as well as good communication~ When the jam is very thick, two towboats of essentially equal power have been used together. They .. te-up bow to bow, and while the propeller wash of one boat loosena and erodes the ice, the second boat holds the first in position. This operation takes a great deal of skill and coordination between the pilots. 3-4. Su..ar:• Each aethod of removing ice j.-described above has its own advantages and disadvantages. The decision aa to which ~thod to use is easy. The difficult problea is to decide if any work is neceaaary. Will the jam go out by itself? Bow great a hazard really exists? !xperi• ence ia helpful for this decision, but ice jaaa are not that oo..an and few people have the opportunity to observe many jama for logical co.pariaon. Thus, advice froa local people familiar with the particular str.ea and its history is invaluable. 3-6