1: \begin{abstract}

2:   We present a robust computational framework for 

3:   advective-diffusive-reactive systems that satisfies 

4:   maximum principles, the non-negative constraint, and 

5:   element-wise species balance property. The proposed 

6:   methodology is valid on general computational grids, 

7:   can handle heterogeneous anisotropic media, and provides 

8:   accurate numerical solutions even for very high P\'eclet 

9:   numbers. 

10:   %

11:   The significant contribution of this paper is to incorporate 

12:   advection (which makes the spatial part of the differential 

13:   operator non-self-adjoint) into the non-negative computational 

14:   framework, and overcome numerical challenges associated 

15:   with advection.

16:   %

17:   We employ low-order mixed finite element formulations 

18:   based on least-squares formalism, and enforce explicit 

19:   constraints on the discrete problem to meet the desired 

20:   properties. 

21:   %

22:   The resulting constrained discrete problem belongs 

23:   to convex quadratic programming for which a unique 

24:   solution exists. Maximum principles and the non-negative 

25:   constraint give rise to bound constraints while element-wise 

26:   species balance gives rise to equality constraints. 

27:   %

28:   The resulting convex quadratic programming problems 

29:   are solved using an interior-point algorithm. Several 

30:   numerical results pertaining to advection-dominated 

31:   problems are presented to illustrate the robustness, 

32:   convergence, and the overall performance of the 

33:   proposed computational framework. 

34: \end{abstract}

35: