AWWA Journal — March 2012
Effect Of Changing Water Quality On Galvanic Coupling
Two galvanic pipe-loop couples (lead–copper and lead– bronze) were exposed to controlled changes in water quality (disinfectant, pH, alkalinity, phosphate) and monitored for changes in lead and copper release. Open circuit potential (OCP) profiles were also measured along the junction of dissimilar metals to determine the extent of the zone affected by galvanic coupling. Grab sampling results showed that changes in water quality caused transient (short-lived) increases in lead and decreases in copper that corresponded to the galvanic action of lead on the other coupled metal. OCP measurements showed that the galvanic effect on corrosion potential can induce a shift of up to 600 mV. The extent of the galvanically affected zone was limited, penetrating no more than a few inches from the juncture along the surface of each pipe. Additional testing confirmed strong effects of external versus direct coupling on the OCP profiles in the galvanically affected zone.
Coupling of two dissimilar surfaces (e.g., lead and copper partial lead service line replacements) results in a predicted and experimentally observed initial increase in the anodic (lead) exchange current, which affects the corrosion process and metal release rates. Recent studies report that galvanic coupling can cause elevated and persistent releases of lead; other studies report that the effect is notable but short-lived. These observations have contributed to uncertainty regarding long-term effects of galvanic couples on lead levels at the tap.
This article focuses on data obtained from galvanically coupled pipe-loop systems that were used to ascertain the effects of varying disinfectants and other water quality changes on lead and copper release (Boyd et al, 2010; 2006). The authors also explored the effects of laboratory testing methods on the galvanic corrosion process and resultant lead and copper releases.
MATERIALS AND METHODS
Pipe-loop setups. Two galvanically coupled pipe-loop setups (lead–copper and lead–bronze) were assembled to represent partial lead service line replacements. The pipe segments were abutted end to end so that the only physical connection between the two pipes occurred at the juncture. At the same testing facility, single-metal pipe-loop setups (new lead, copper, bronze, and aged lead) were assembled and operated similarly. All setups were designed to control the targeted water quality, quantify lead and copper release by grab sampling, and measure electrical surface potentials.
The study was divided into four test sequences, with each test sequence lasting approximately three months. A test sequence was designed to represent a cycle of changing disinfectants overlaid by a permutation of background water quality. Test 1 served as the control. For test 2, the cycle of changing disinfectants was overlaid with an alkalinity increase to 140 mg/L as calcium carbonate. For test 3, the pH was decreased to 7.0–7.5. For test 4, phosphate was added to produce a residual of 1 mg/L as PO4. A fresh supply of test water was prepared once per week, water quality adjustments were made daily, and samples were collected and analyzed for total lead and copper three to five times per week. OCPs were measured once per week along the interior surface of each coupled pipe specimen.
Immersed coupons. Following completion of the pipeloop tests, additional bench-scale testing was conducted to understand the effects of galvanic coupling configurations on OCPs. Coupons were connected externally or directly. For the externally coupled coupons, a copper wire with clamps connected the independent coupons. For the directly coupled coupon, a copper pipe was sectioned, and one end was dipped in molten lead.
RESULTS AND DISCISSION
Sampling results. Figure 1 compares lead levels released from lead–copper galvanic coupling (loop B3) and singlemetal unpassivated lead pipe (loop A1). Lead levels increased slightly when the galvanic couple was initially exposed to a changed water quality in test 2 and considerably more prominently in test 3. A transient increase of lead also occurred at the beginning of test 4. A change in water quality induced a transient increase in lead release; however, this effect was short lived. After acclimation, lead leaching from the galvanic pipe loop approached levels comparable to the single-metal pipe loop.
Copper levels from the same lead–copper galvanic pipe loop were similarly compared with the single-metal copper pipe loop. Observations were consistent with the fact that the acceleration of lead oxidation in the galvanic couple was accompanied by a temporary increase of the cathodic protection of the copper, which resulted in a decrease of copper release. The copper concentrations subsequently increased to their initial levels. This concerted behavior of lead and copper concentrations indicated that galvanic acceleration of lead release had a transient nature. Similar trends were observed in lead and copper levels collected from the lead–bronze pipe loop.
OCP measurements. OCP data were collected from the directly connected lead–copper galvanic pipe-loop couple. In all cases, the magnitude of the potential shift was substantial (up to 600 mV). Results provided evidence that the length of the zone in which the galvanic coupling exerted its influence was quite limited, penetrating no more than a few inches from the juncture. Similar results were observed for the lead–bronze couple.
Surface potential measurements were plotted as a function of distance for separated coupons and coupons connected with an external wire. Once the external contact was established, the highly negative potential of the entire lead coupon shifted in an anodic (positive) direction, whereas the more positive potential of the entire copper coupon shifted in a cathodic (negative) direction. These shifts demonstrated that lead corrosion was accelerated and copper corrosion was retarded. The results showed the importance of the external versus direct coupling configuration, its effect on the galvanic corrosion process, and its influence on lead and copper release rates.
Lead and copper samples collected from recirculating pipe loops containing lead–copper (and lead–bronze) galvanic couplings indicated that changes in water quality caused an initial, transient increase in lead and a corresponding decrease in copper leaching. The data also demonstrated that the effect was short-lived.
OCP measurements were collected along the interior surface of each of the galvanically coupled pipes. Measurements showed that the OCP underwent consistent changes as a function of distance from the point of contact. Experimental data further demonstrated that the extent of the observed galvanic coupling effect was limited, penetrating no more than a few inches from the juncture.
Immersed coupon experiments demonstrated significant differences in surface potential profiles for directly versus externally coupled galvanic configurations. For the external configuration, the profiles shifted toward each other, thus accelerating lead corrosion and retarding copper corrosion. More research is needed to understand the relative galvanic effects of direct versus external metal coupling configurations.
Boyd, G.R.; McFadden, M.S.; Reiber, S.H.; Sandvig, A.M.; Korshin, G.V.; Giani, R.; & Frenkel, A.I., 2010. Effect of Changing Disinfectants on Distribution System Lead and Copper Release: Part 2—Final Report. Water Research Foundation, Denver.
Boyd, G.R.; Dewis, K.M.; Sandvig, A.M.; Kirmeyer, G.J.; Reiber, S.H.; & Korshin, G.V., 2006. Effect of Changing Disinfectants on Distribution System Lead and Copper Release, Part 1—Literature Review. Awwa Research Foundation, Denver.
Corresponding author: Glen R. Boyd is a senior associate with The Cadmus Group, 1411 Fourth Ave., Ste. 1106, Seattle, WA 98101; email@example.com.
A full report of this project, Effect of Changing Disinfectants on Distribution System Lead and Copper Release: Part 2— Research Results (3107), is available for free to Water Research Foundation subscribers by logging on to www.waterrf.org.