Uncontrollable leakage of pancreatic secretion causes a metabolic acidosis due to the loss of strong cations.

Pancreaticoenteric fistula as a cause of normal anion gap metabolic acidosis

This is mainly due to the highly bicarbonate-rich fluid issuing from the pancreas. Pancreatic secretions are essentially iso-osmolar with the body fluid, and contain about the same amount of sodium and potassium as the plasma. Not so for the chloride and bicarbonate.

Under the influence of VIP (Vasoactive Intestinal Peptide), the pancreatic CTFR (Cystic Fibrosis Transmembrane conductance Regulator) constantly secretes chloride into the lumen of the pancreatic duct; this secreted chloride serves as the fodder for a chloride-bicarbonate anion exchanger, which enriches the lumen with bicarbonate as it dechlorinates the pancreatic secretions. The sodium and potassium equilibrate easily across the leaky membranes.

uncontrolled pancreatic secretion as a cause of metabolic acidosis

Not only that, but it seems that the higher the rate of pancreatic secretion, the more bicarbonate and less chloride is secreted- at least in the disembodied pancreas of the domestic cat.

graph chloride concentration in pancreatic secretions decreases as rate of secretion increases

This graph is loosely adopted  from the 1968 study by Case Harper and Scratcherd. Notice how even initially the concentration of bicarbonate in the fluid is very high; and the greater the rate of secretion, the more bicarbonate there is.

Of course, the bicarbonate is a dependent variable in Stewarts equation; the more relevant feature is the rate of chloride secretion, which decreases. The sodium excretion rate, however, remains stable.

This, of course, means a retention of one strong ion (chloride) and an excretion of another (sodium).

The decrease in strong ion difference leads to a metabolic acidosis with a raised chloride.

It is known that we humans secrete about 1.5L of pancreatic juice per day – on average about 60ml/hr. One can calculate that this leads to the secretion of 217.5 mmol of Na+ each day.

Given that there would be no problem if the chloride was secreted at an equivalent rate, and calculating for a minimal chloride concentration of pancreatic secretions (40mmol/L) we can calculate that for every litre of pancreatic secretion, 60mmol of chloride is disproportionally retained (or 90mmol each day).

Now, this electrolyte leak is corrected by the time the pancreatic juices mixed with food reach the end of the intestine. The small intestine and colon play a major role in reclaiming the sodum and chloride, as the bicarbonate is converted into CO2 and water. Indeed, the osmolarity of normal stool water is essentially the same as that of plasma, but by that stage the 7 or so litres of fluid which traffics throught the gut on a daily basis will turn into about 150ml of fluid, excreted along with the solid stool. 

However, let us think of what might happen if the pancreatic secretions are diverted, so there is no opportunity for electrolyte reabsorption. 

A gradual decrease in the strong ion difference will result.

Metabolic acidosis ensues.

References

Fawcett, A. N., and C. A. Newman. "The effect of prolonged pancreatic secretion on blood acid-base balance in the conscious dog." Gut 12.11 (1971): 918-921.

Nghiem, Dai D., Thomas A. Gonwa, and Robert J. Corry. "Metabolic effects of urinary diversion of exocrine secretions in pancreatic transplantation." Transplantation 43.1 (1987): 70-72.

Case, R. Mo, A. A. Harper, and T. Scratcherd. "Water and electrolyte secretion by the perfused pancreas of the cat." The Journal of physiology 196.1 (1968): 133-149.

Eovaldi, Benjamin, and Claude Zanetti. "Non–Anion Gap Metabolic Acidosis in a Patient With a Pancreaticopleural Fistula." JAOA: Journal of the American Osteopathic Association 111.5 (2011): 344-345.

Fordtran JS, Rector FC, Jr, Ewton MF, Soter N, Kinney J. Permeability characteristics of the human small intestine. J. Clin. Invest. 1965;44:1935–1944.

Field, Michael. "Intestinal ion transport and the pathophysiology of diarrhea."Journal of Clinical Investigation 111.7 (2003): 931-943.